subsurface/core/planner.c

1230 lines
40 KiB
C
Raw Normal View History

// SPDX-License-Identifier: GPL-2.0
First stab at simplistic dive planning This comes with absolutely no gui - so the plan literally needs to be compiled into Subsurface. Not exactly a feature, but this allowed me to focus on the planning part instead of spending time on tedious UI work. A new menu "Planner" with entry "Test Planner" calls into the hard-coded function in planner.c. There a simple dive plan can be constructed with calls to plan_add_segment(&diveplan, duration, depth at the end, fO2, pO2) Calling plan(&diveplan) does the deco calculations and creates deco stops that keep us below the ceiling (with the GFlow/high values currently configured). The stop levels used are defined at the top of planner.c in the stoplevels array - there is no need to do the traditional multiples of 3m or anything like that. The dive including the ascents and deco stops all the way to the surface is completed and then added as simulated dive to the end of the divelist (I guess we could automatically select it later) and can be viewed. This is crude but shows the direction we can go with this. Envision a nice UI that allows you to simply enter the segments and pick the desired stops. What is missing is the ability to give the algorithm additional gases that it can use during the deco phase - right now it simply keeps using the last gas used in the diveplan. All that said, there are clear bugs here - and sadly they seem to be in the deco calculations, as with the example given the ceiling that is calculated makes no sense. When displayed in smooth mode it has very strange jumps up and down that I wouldn't expect. For example with GF 35/75 (the default) the deco ceiling when looking at the simulated dive jumps from 16m back up to 13m around 14:10 into the dive. That seems very odd. Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2013-01-05 07:11:42 +00:00
/* planner.c
*
* code that allows us to plan future dives
*
* (c) Dirk Hohndel 2013
*/
#include <assert.h>
#include <unistd.h>
#include <ctype.h>
#include <string.h>
#include "ssrf.h"
#include "divelist.h"
#include "subsurface-string.h"
#include "deco.h"
#include "errorhelper.h"
#include "planner.h"
#include "gettext.h"
#include "libdivecomputer/parser.h"
#include "qthelper.h"
#include "version.h"
First stab at simplistic dive planning This comes with absolutely no gui - so the plan literally needs to be compiled into Subsurface. Not exactly a feature, but this allowed me to focus on the planning part instead of spending time on tedious UI work. A new menu "Planner" with entry "Test Planner" calls into the hard-coded function in planner.c. There a simple dive plan can be constructed with calls to plan_add_segment(&diveplan, duration, depth at the end, fO2, pO2) Calling plan(&diveplan) does the deco calculations and creates deco stops that keep us below the ceiling (with the GFlow/high values currently configured). The stop levels used are defined at the top of planner.c in the stoplevels array - there is no need to do the traditional multiples of 3m or anything like that. The dive including the ascents and deco stops all the way to the surface is completed and then added as simulated dive to the end of the divelist (I guess we could automatically select it later) and can be viewed. This is crude but shows the direction we can go with this. Envision a nice UI that allows you to simply enter the segments and pick the desired stops. What is missing is the ability to give the algorithm additional gases that it can use during the deco phase - right now it simply keeps using the last gas used in the diveplan. All that said, there are clear bugs here - and sadly they seem to be in the deco calculations, as with the example given the ceiling that is calculated makes no sense. When displayed in smooth mode it has very strange jumps up and down that I wouldn't expect. For example with GF 35/75 (the default) the deco ceiling when looking at the simulated dive jumps from 16m back up to 13m around 14:10 into the dive. That seems very odd. Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2013-01-05 07:11:42 +00:00
#define TIMESTEP 2 /* second */
static int decostoplevels_metric[] = { 0, 3000, 6000, 9000, 12000, 15000, 18000, 21000, 24000, 27000,
30000, 33000, 36000, 39000, 42000, 45000, 48000, 51000, 54000, 57000,
60000, 63000, 66000, 69000, 72000, 75000, 78000, 81000, 84000, 87000,
90000, 100000, 110000, 120000, 130000, 140000, 150000, 160000, 170000,
180000, 190000, 200000, 220000, 240000, 260000, 280000, 300000,
320000, 340000, 360000, 380000 };
static int decostoplevels_imperial[] = { 0, 3048, 6096, 9144, 12192, 15240, 18288, 21336, 24384, 27432,
30480, 33528, 36576, 39624, 42672, 45720, 48768, 51816, 54864, 57912,
60960, 64008, 67056, 70104, 73152, 76200, 79248, 82296, 85344, 88392,
91440, 101600, 111760, 121920, 132080, 142240, 152400, 162560, 172720,
182880, 193040, 203200, 223520, 243840, 264160, 284480, 304800,
325120, 345440, 365760, 386080 };
#if DEBUG_PLAN
void dump_plan(struct diveplan *diveplan)
{
struct divedatapoint *dp;
struct tm tm;
if (!diveplan) {
printf("Diveplan NULL\n");
return;
}
utc_mkdate(diveplan->when, &tm);
printf("\nDiveplan @ %04d-%02d-%02d %02d:%02d:%02d (surfpres %dmbar):\n",
tm.tm_year, tm.tm_mon + 1, tm.tm_mday,
tm.tm_hour, tm.tm_min, tm.tm_sec,
diveplan->surface_pressure);
dp = diveplan->dp;
while (dp) {
printf("\t%3u:%02u: %6dmm cylid: %2d setpoint: %d\n", FRACTION(dp->time, 60), dp->depth, dp->cylinderid, dp->setpoint);
dp = dp->next;
}
}
#endif
bool diveplan_empty(struct diveplan *diveplan)
{
struct divedatapoint *dp;
if (!diveplan || !diveplan->dp)
return true;
dp = diveplan->dp;
while (dp) {
if (dp->time)
return false;
dp = dp->next;
}
return true;
}
/* get the cylinder index at a certain time during the dive */
int get_cylinderid_at_time(struct dive *dive, struct divecomputer *dc, duration_t time)
{
// we start with the first cylinder unless an event tells us otherwise
int cylinder_idx = 0;
struct event *event = dc->events;
while (event && event->time.seconds <= time.seconds) {
if (!strcmp(event->name, "gaschange"))
cylinder_idx = get_cylinder_index(dive, event);
event = event->next;
}
return cylinder_idx;
}
int get_gasidx(struct dive *dive, struct gasmix mix)
{
return find_best_gasmix_match(mix, &dive->cylinders);
}
static void interpolate_transition(struct deco_state *ds, struct dive *dive, duration_t t0, duration_t t1, depth_t d0, depth_t d1, struct gasmix gasmix, o2pressure_t po2, enum divemode_t divemode)
Improvement for various bits of the planner Rewrite of the actual planner logic. Now ascend to the next potential stop depth. There the state is cached and we try to ascend to the next stop depth. If we hit the ceiling while doing that we go back to the cached state and wait there for a minute. Then we try again. Then loop. Converted all depth related variables from unsigned int to int. During planning, in a time step the current depth can temporarily be negative and comparisons of a negative int with an unsigned it have not the result I expected ( (int) -2 < (unsigned int) 3 turns out to be false). And we don’t really need the 32nd bit that unsigned buys us for depths. Deco stops are now shown in the same table as manually entered stops in boldface (I removed the second table to save screen estate). The gas shown in the table is still misleading as it means the gas used on the segment leading up to that event. The update of the profile only works partially upon changes in the list of available gases. Treatment of various gases is basically there but needs some more love. The ascent velocity is now provided by a function that takes the current depth as argument. Currently it always returns 10m/min but that will later be variable (and hopefully user configurable). The profile is not redrawn while deco is computed (avoiding an infinite recursion). The table got a new column for the duration of a segment while the old “duration” column was renamed “Runtime” to reflect what it actually shows. Currently, only the run time but not the duration are editable. All deco gases are used from the depth where their pO2 is 1.4bar. This should become more flexible. Calculation of the pressure drop in cylinders without configured volumes is suppressed. This solves a problem with the planner crashing when saving a dive where not all cylinders had been manually given a volume. [Short rant break: Treating 0/0 as air bites back at so many places. E.g. Cylinder data is initialized with memsetting the whole structures to 0. Then later suddenly this totally unconfigured cylinder is being treated as it would contain air. Maybe at some point this was a feature. But it lead to a naughty bug which took me over an hour to resolve. We should seriously reconsider this choice and better move to 209/0 being air if changing this everywhere is not too much trouble] Signed-off-by: Robert C. Helling <helling@atdotde.de> Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2014-04-17 08:54:55 +00:00
{
int32_t j;
Improvement for various bits of the planner Rewrite of the actual planner logic. Now ascend to the next potential stop depth. There the state is cached and we try to ascend to the next stop depth. If we hit the ceiling while doing that we go back to the cached state and wait there for a minute. Then we try again. Then loop. Converted all depth related variables from unsigned int to int. During planning, in a time step the current depth can temporarily be negative and comparisons of a negative int with an unsigned it have not the result I expected ( (int) -2 < (unsigned int) 3 turns out to be false). And we don’t really need the 32nd bit that unsigned buys us for depths. Deco stops are now shown in the same table as manually entered stops in boldface (I removed the second table to save screen estate). The gas shown in the table is still misleading as it means the gas used on the segment leading up to that event. The update of the profile only works partially upon changes in the list of available gases. Treatment of various gases is basically there but needs some more love. The ascent velocity is now provided by a function that takes the current depth as argument. Currently it always returns 10m/min but that will later be variable (and hopefully user configurable). The profile is not redrawn while deco is computed (avoiding an infinite recursion). The table got a new column for the duration of a segment while the old “duration” column was renamed “Runtime” to reflect what it actually shows. Currently, only the run time but not the duration are editable. All deco gases are used from the depth where their pO2 is 1.4bar. This should become more flexible. Calculation of the pressure drop in cylinders without configured volumes is suppressed. This solves a problem with the planner crashing when saving a dive where not all cylinders had been manually given a volume. [Short rant break: Treating 0/0 as air bites back at so many places. E.g. Cylinder data is initialized with memsetting the whole structures to 0. Then later suddenly this totally unconfigured cylinder is being treated as it would contain air. Maybe at some point this was a feature. But it lead to a naughty bug which took me over an hour to resolve. We should seriously reconsider this choice and better move to 209/0 being air if changing this everywhere is not too much trouble] Signed-off-by: Robert C. Helling <helling@atdotde.de> Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2014-04-17 08:54:55 +00:00
for (j = t0.seconds; j < t1.seconds; j++) {
int depth = interpolate(d0.mm, d1.mm, j - t0.seconds, t1.seconds - t0.seconds);
add_segment(ds, depth_to_bar(depth, dive), gasmix, 1, po2.mbar, divemode, prefs.bottomsac);
Improvement for various bits of the planner Rewrite of the actual planner logic. Now ascend to the next potential stop depth. There the state is cached and we try to ascend to the next stop depth. If we hit the ceiling while doing that we go back to the cached state and wait there for a minute. Then we try again. Then loop. Converted all depth related variables from unsigned int to int. During planning, in a time step the current depth can temporarily be negative and comparisons of a negative int with an unsigned it have not the result I expected ( (int) -2 < (unsigned int) 3 turns out to be false). And we don’t really need the 32nd bit that unsigned buys us for depths. Deco stops are now shown in the same table as manually entered stops in boldface (I removed the second table to save screen estate). The gas shown in the table is still misleading as it means the gas used on the segment leading up to that event. The update of the profile only works partially upon changes in the list of available gases. Treatment of various gases is basically there but needs some more love. The ascent velocity is now provided by a function that takes the current depth as argument. Currently it always returns 10m/min but that will later be variable (and hopefully user configurable). The profile is not redrawn while deco is computed (avoiding an infinite recursion). The table got a new column for the duration of a segment while the old “duration” column was renamed “Runtime” to reflect what it actually shows. Currently, only the run time but not the duration are editable. All deco gases are used from the depth where their pO2 is 1.4bar. This should become more flexible. Calculation of the pressure drop in cylinders without configured volumes is suppressed. This solves a problem with the planner crashing when saving a dive where not all cylinders had been manually given a volume. [Short rant break: Treating 0/0 as air bites back at so many places. E.g. Cylinder data is initialized with memsetting the whole structures to 0. Then later suddenly this totally unconfigured cylinder is being treated as it would contain air. Maybe at some point this was a feature. But it lead to a naughty bug which took me over an hour to resolve. We should seriously reconsider this choice and better move to 209/0 being air if changing this everywhere is not too much trouble] Signed-off-by: Robert C. Helling <helling@atdotde.de> Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2014-04-17 08:54:55 +00:00
}
if (d1.mm > d0.mm)
calc_crushing_pressure(ds, depth_to_bar(d1.mm, dive));
Improvement for various bits of the planner Rewrite of the actual planner logic. Now ascend to the next potential stop depth. There the state is cached and we try to ascend to the next stop depth. If we hit the ceiling while doing that we go back to the cached state and wait there for a minute. Then we try again. Then loop. Converted all depth related variables from unsigned int to int. During planning, in a time step the current depth can temporarily be negative and comparisons of a negative int with an unsigned it have not the result I expected ( (int) -2 < (unsigned int) 3 turns out to be false). And we don’t really need the 32nd bit that unsigned buys us for depths. Deco stops are now shown in the same table as manually entered stops in boldface (I removed the second table to save screen estate). The gas shown in the table is still misleading as it means the gas used on the segment leading up to that event. The update of the profile only works partially upon changes in the list of available gases. Treatment of various gases is basically there but needs some more love. The ascent velocity is now provided by a function that takes the current depth as argument. Currently it always returns 10m/min but that will later be variable (and hopefully user configurable). The profile is not redrawn while deco is computed (avoiding an infinite recursion). The table got a new column for the duration of a segment while the old “duration” column was renamed “Runtime” to reflect what it actually shows. Currently, only the run time but not the duration are editable. All deco gases are used from the depth where their pO2 is 1.4bar. This should become more flexible. Calculation of the pressure drop in cylinders without configured volumes is suppressed. This solves a problem with the planner crashing when saving a dive where not all cylinders had been manually given a volume. [Short rant break: Treating 0/0 as air bites back at so many places. E.g. Cylinder data is initialized with memsetting the whole structures to 0. Then later suddenly this totally unconfigured cylinder is being treated as it would contain air. Maybe at some point this was a feature. But it lead to a naughty bug which took me over an hour to resolve. We should seriously reconsider this choice and better move to 209/0 being air if changing this everywhere is not too much trouble] Signed-off-by: Robert C. Helling <helling@atdotde.de> Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2014-04-17 08:54:55 +00:00
}
First stab at simplistic dive planning This comes with absolutely no gui - so the plan literally needs to be compiled into Subsurface. Not exactly a feature, but this allowed me to focus on the planning part instead of spending time on tedious UI work. A new menu "Planner" with entry "Test Planner" calls into the hard-coded function in planner.c. There a simple dive plan can be constructed with calls to plan_add_segment(&diveplan, duration, depth at the end, fO2, pO2) Calling plan(&diveplan) does the deco calculations and creates deco stops that keep us below the ceiling (with the GFlow/high values currently configured). The stop levels used are defined at the top of planner.c in the stoplevels array - there is no need to do the traditional multiples of 3m or anything like that. The dive including the ascents and deco stops all the way to the surface is completed and then added as simulated dive to the end of the divelist (I guess we could automatically select it later) and can be viewed. This is crude but shows the direction we can go with this. Envision a nice UI that allows you to simply enter the segments and pick the desired stops. What is missing is the ability to give the algorithm additional gases that it can use during the deco phase - right now it simply keeps using the last gas used in the diveplan. All that said, there are clear bugs here - and sadly they seem to be in the deco calculations, as with the example given the ceiling that is calculated makes no sense. When displayed in smooth mode it has very strange jumps up and down that I wouldn't expect. For example with GF 35/75 (the default) the deco ceiling when looking at the simulated dive jumps from 16m back up to 13m around 14:10 into the dive. That seems very odd. Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2013-01-05 07:11:42 +00:00
/* returns the tissue tolerance at the end of this (partial) dive */
static int tissue_at_end(struct deco_state *ds, struct dive *dive, struct deco_state **cached_datap)
First stab at simplistic dive planning This comes with absolutely no gui - so the plan literally needs to be compiled into Subsurface. Not exactly a feature, but this allowed me to focus on the planning part instead of spending time on tedious UI work. A new menu "Planner" with entry "Test Planner" calls into the hard-coded function in planner.c. There a simple dive plan can be constructed with calls to plan_add_segment(&diveplan, duration, depth at the end, fO2, pO2) Calling plan(&diveplan) does the deco calculations and creates deco stops that keep us below the ceiling (with the GFlow/high values currently configured). The stop levels used are defined at the top of planner.c in the stoplevels array - there is no need to do the traditional multiples of 3m or anything like that. The dive including the ascents and deco stops all the way to the surface is completed and then added as simulated dive to the end of the divelist (I guess we could automatically select it later) and can be viewed. This is crude but shows the direction we can go with this. Envision a nice UI that allows you to simply enter the segments and pick the desired stops. What is missing is the ability to give the algorithm additional gases that it can use during the deco phase - right now it simply keeps using the last gas used in the diveplan. All that said, there are clear bugs here - and sadly they seem to be in the deco calculations, as with the example given the ceiling that is calculated makes no sense. When displayed in smooth mode it has very strange jumps up and down that I wouldn't expect. For example with GF 35/75 (the default) the deco ceiling when looking at the simulated dive jumps from 16m back up to 13m around 14:10 into the dive. That seems very odd. Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2013-01-05 07:11:42 +00:00
{
struct divecomputer *dc;
struct sample *sample, *psample;
int i;
depth_t lastdepth = {};
duration_t t0 = {}, t1 = {};
struct gasmix gas;
int surface_interval = 0;
First stab at simplistic dive planning This comes with absolutely no gui - so the plan literally needs to be compiled into Subsurface. Not exactly a feature, but this allowed me to focus on the planning part instead of spending time on tedious UI work. A new menu "Planner" with entry "Test Planner" calls into the hard-coded function in planner.c. There a simple dive plan can be constructed with calls to plan_add_segment(&diveplan, duration, depth at the end, fO2, pO2) Calling plan(&diveplan) does the deco calculations and creates deco stops that keep us below the ceiling (with the GFlow/high values currently configured). The stop levels used are defined at the top of planner.c in the stoplevels array - there is no need to do the traditional multiples of 3m or anything like that. The dive including the ascents and deco stops all the way to the surface is completed and then added as simulated dive to the end of the divelist (I guess we could automatically select it later) and can be viewed. This is crude but shows the direction we can go with this. Envision a nice UI that allows you to simply enter the segments and pick the desired stops. What is missing is the ability to give the algorithm additional gases that it can use during the deco phase - right now it simply keeps using the last gas used in the diveplan. All that said, there are clear bugs here - and sadly they seem to be in the deco calculations, as with the example given the ceiling that is calculated makes no sense. When displayed in smooth mode it has very strange jumps up and down that I wouldn't expect. For example with GF 35/75 (the default) the deco ceiling when looking at the simulated dive jumps from 16m back up to 13m around 14:10 into the dive. That seems very odd. Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2013-01-05 07:11:42 +00:00
if (!dive)
return 0;
if (*cached_datap) {
restore_deco_state(*cached_datap, ds, true);
} else {
surface_interval = init_decompression(ds, dive);
cache_deco_state(ds, cached_datap);
}
First stab at simplistic dive planning This comes with absolutely no gui - so the plan literally needs to be compiled into Subsurface. Not exactly a feature, but this allowed me to focus on the planning part instead of spending time on tedious UI work. A new menu "Planner" with entry "Test Planner" calls into the hard-coded function in planner.c. There a simple dive plan can be constructed with calls to plan_add_segment(&diveplan, duration, depth at the end, fO2, pO2) Calling plan(&diveplan) does the deco calculations and creates deco stops that keep us below the ceiling (with the GFlow/high values currently configured). The stop levels used are defined at the top of planner.c in the stoplevels array - there is no need to do the traditional multiples of 3m or anything like that. The dive including the ascents and deco stops all the way to the surface is completed and then added as simulated dive to the end of the divelist (I guess we could automatically select it later) and can be viewed. This is crude but shows the direction we can go with this. Envision a nice UI that allows you to simply enter the segments and pick the desired stops. What is missing is the ability to give the algorithm additional gases that it can use during the deco phase - right now it simply keeps using the last gas used in the diveplan. All that said, there are clear bugs here - and sadly they seem to be in the deco calculations, as with the example given the ceiling that is calculated makes no sense. When displayed in smooth mode it has very strange jumps up and down that I wouldn't expect. For example with GF 35/75 (the default) the deco ceiling when looking at the simulated dive jumps from 16m back up to 13m around 14:10 into the dive. That seems very odd. Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2013-01-05 07:11:42 +00:00
dc = &dive->dc;
if (!dc->samples)
return 0;
First stab at simplistic dive planning This comes with absolutely no gui - so the plan literally needs to be compiled into Subsurface. Not exactly a feature, but this allowed me to focus on the planning part instead of spending time on tedious UI work. A new menu "Planner" with entry "Test Planner" calls into the hard-coded function in planner.c. There a simple dive plan can be constructed with calls to plan_add_segment(&diveplan, duration, depth at the end, fO2, pO2) Calling plan(&diveplan) does the deco calculations and creates deco stops that keep us below the ceiling (with the GFlow/high values currently configured). The stop levels used are defined at the top of planner.c in the stoplevels array - there is no need to do the traditional multiples of 3m or anything like that. The dive including the ascents and deco stops all the way to the surface is completed and then added as simulated dive to the end of the divelist (I guess we could automatically select it later) and can be viewed. This is crude but shows the direction we can go with this. Envision a nice UI that allows you to simply enter the segments and pick the desired stops. What is missing is the ability to give the algorithm additional gases that it can use during the deco phase - right now it simply keeps using the last gas used in the diveplan. All that said, there are clear bugs here - and sadly they seem to be in the deco calculations, as with the example given the ceiling that is calculated makes no sense. When displayed in smooth mode it has very strange jumps up and down that I wouldn't expect. For example with GF 35/75 (the default) the deco ceiling when looking at the simulated dive jumps from 16m back up to 13m around 14:10 into the dive. That seems very odd. Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2013-01-05 07:11:42 +00:00
psample = sample = dc->sample;
const struct event *evdm = NULL;
enum divemode_t divemode = UNDEF_COMP_TYPE;
First stab at simplistic dive planning This comes with absolutely no gui - so the plan literally needs to be compiled into Subsurface. Not exactly a feature, but this allowed me to focus on the planning part instead of spending time on tedious UI work. A new menu "Planner" with entry "Test Planner" calls into the hard-coded function in planner.c. There a simple dive plan can be constructed with calls to plan_add_segment(&diveplan, duration, depth at the end, fO2, pO2) Calling plan(&diveplan) does the deco calculations and creates deco stops that keep us below the ceiling (with the GFlow/high values currently configured). The stop levels used are defined at the top of planner.c in the stoplevels array - there is no need to do the traditional multiples of 3m or anything like that. The dive including the ascents and deco stops all the way to the surface is completed and then added as simulated dive to the end of the divelist (I guess we could automatically select it later) and can be viewed. This is crude but shows the direction we can go with this. Envision a nice UI that allows you to simply enter the segments and pick the desired stops. What is missing is the ability to give the algorithm additional gases that it can use during the deco phase - right now it simply keeps using the last gas used in the diveplan. All that said, there are clear bugs here - and sadly they seem to be in the deco calculations, as with the example given the ceiling that is calculated makes no sense. When displayed in smooth mode it has very strange jumps up and down that I wouldn't expect. For example with GF 35/75 (the default) the deco ceiling when looking at the simulated dive jumps from 16m back up to 13m around 14:10 into the dive. That seems very odd. Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2013-01-05 07:11:42 +00:00
for (i = 0; i < dc->samples; i++, sample++) {
o2pressure_t setpoint;
if (i)
setpoint = sample[-1].setpoint;
else
setpoint = sample[0].setpoint;
t1 = sample->time;
gas = get_gasmix_at_time(dive, dc, t0);
if (i > 0)
lastdepth = psample->depth;
/* The ceiling in the deeper portion of a multilevel dive is sometimes critical for the VPM-B
* Boyle's law compensation. We should check the ceiling prior to ascending during the bottom
* portion of the dive. The maximum ceiling might be reached while ascending, but testing indicates
* that it is only marginally deeper than the ceiling at the start of ascent.
* Do not set the first_ceiling_pressure variable (used for the Boyle's law compensation calculation)
* at this stage, because it would interfere with calculating the ceiling at the end of the bottom
* portion of the dive.
* Remember the value for later.
*/
if ((decoMode() == VPMB) && (lastdepth.mm > sample->depth.mm)) {
pressure_t ceiling_pressure;
nuclear_regeneration(ds, t0.seconds);
vpmb_start_gradient(ds);
ceiling_pressure.mbar = depth_to_mbar(deco_allowed_depth(tissue_tolerance_calc(ds, dive,
depth_to_bar(lastdepth.mm, dive)),
dive->surface_pressure.mbar / 1000.0,
dive,
1),
dive);
if (ceiling_pressure.mbar > ds->max_bottom_ceiling_pressure.mbar)
ds->max_bottom_ceiling_pressure.mbar = ceiling_pressure.mbar;
}
divemode = get_current_divemode(&dive->dc, t0.seconds + 1, &evdm, &divemode);
interpolate_transition(ds, dive, t0, t1, lastdepth, sample->depth, gas, setpoint, divemode);
First stab at simplistic dive planning This comes with absolutely no gui - so the plan literally needs to be compiled into Subsurface. Not exactly a feature, but this allowed me to focus on the planning part instead of spending time on tedious UI work. A new menu "Planner" with entry "Test Planner" calls into the hard-coded function in planner.c. There a simple dive plan can be constructed with calls to plan_add_segment(&diveplan, duration, depth at the end, fO2, pO2) Calling plan(&diveplan) does the deco calculations and creates deco stops that keep us below the ceiling (with the GFlow/high values currently configured). The stop levels used are defined at the top of planner.c in the stoplevels array - there is no need to do the traditional multiples of 3m or anything like that. The dive including the ascents and deco stops all the way to the surface is completed and then added as simulated dive to the end of the divelist (I guess we could automatically select it later) and can be viewed. This is crude but shows the direction we can go with this. Envision a nice UI that allows you to simply enter the segments and pick the desired stops. What is missing is the ability to give the algorithm additional gases that it can use during the deco phase - right now it simply keeps using the last gas used in the diveplan. All that said, there are clear bugs here - and sadly they seem to be in the deco calculations, as with the example given the ceiling that is calculated makes no sense. When displayed in smooth mode it has very strange jumps up and down that I wouldn't expect. For example with GF 35/75 (the default) the deco ceiling when looking at the simulated dive jumps from 16m back up to 13m around 14:10 into the dive. That seems very odd. Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2013-01-05 07:11:42 +00:00
psample = sample;
t0 = t1;
}
return surface_interval;
First stab at simplistic dive planning This comes with absolutely no gui - so the plan literally needs to be compiled into Subsurface. Not exactly a feature, but this allowed me to focus on the planning part instead of spending time on tedious UI work. A new menu "Planner" with entry "Test Planner" calls into the hard-coded function in planner.c. There a simple dive plan can be constructed with calls to plan_add_segment(&diveplan, duration, depth at the end, fO2, pO2) Calling plan(&diveplan) does the deco calculations and creates deco stops that keep us below the ceiling (with the GFlow/high values currently configured). The stop levels used are defined at the top of planner.c in the stoplevels array - there is no need to do the traditional multiples of 3m or anything like that. The dive including the ascents and deco stops all the way to the surface is completed and then added as simulated dive to the end of the divelist (I guess we could automatically select it later) and can be viewed. This is crude but shows the direction we can go with this. Envision a nice UI that allows you to simply enter the segments and pick the desired stops. What is missing is the ability to give the algorithm additional gases that it can use during the deco phase - right now it simply keeps using the last gas used in the diveplan. All that said, there are clear bugs here - and sadly they seem to be in the deco calculations, as with the example given the ceiling that is calculated makes no sense. When displayed in smooth mode it has very strange jumps up and down that I wouldn't expect. For example with GF 35/75 (the default) the deco ceiling when looking at the simulated dive jumps from 16m back up to 13m around 14:10 into the dive. That seems very odd. Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2013-01-05 07:11:42 +00:00
}
/* if a default cylinder is set, use that */
void fill_default_cylinder(const struct dive *dive, cylinder_t *cyl)
{
const char *cyl_name = prefs.default_cylinder;
struct tank_info_t *ti = tank_info;
pressure_t pO2 = {.mbar = 1600};
if (!cyl_name)
return;
while (ti->name != NULL && ti < tank_info + MAX_TANK_INFO) {
if (strcmp(ti->name, cyl_name) == 0)
break;
ti++;
}
if (ti->name == NULL)
/* didn't find it */
return;
cyl->type.description = strdup(ti->name);
if (ti->ml) {
cyl->type.size.mliter = ti->ml;
cyl->type.workingpressure.mbar = ti->bar * 1000;
} else {
cyl->type.workingpressure.mbar = psi_to_mbar(ti->psi);
if (ti->psi)
cyl->type.size.mliter = lrint(cuft_to_l(ti->cuft) * 1000 / bar_to_atm(psi_to_bar(ti->psi)));
}
// MOD of air
cyl->depth = gas_mod(cyl->gasmix, pO2, dive, 1);
}
/* calculate the new end pressure of the cylinder, based on its current end pressure and the
* latest segment. */
static void update_cylinder_pressure(struct dive *d, int old_depth, int new_depth, int duration, int sac, cylinder_t *cyl, bool in_deco, enum divemode_t divemode)
{
volume_t gas_used;
pressure_t delta_p;
depth_t mean_depth;
int factor = 1000;
if (divemode == PSCR)
factor = prefs.pscr_ratio;
if (!cyl)
return;
mean_depth.mm = (old_depth + new_depth) / 2;
gas_used.mliter = lrint(depth_to_atm(mean_depth.mm, d) * sac / 60 * duration * factor / 1000);
cyl->gas_used.mliter += gas_used.mliter;
if (in_deco)
cyl->deco_gas_used.mliter += gas_used.mliter;
if (cyl->type.size.mliter) {
delta_p.mbar = lrint(gas_used.mliter * 1000.0 / cyl->type.size.mliter * gas_compressibility_factor(cyl->gasmix, cyl->end.mbar / 1000.0));
cyl->end.mbar -= delta_p.mbar;
}
First stab at simplistic dive planning This comes with absolutely no gui - so the plan literally needs to be compiled into Subsurface. Not exactly a feature, but this allowed me to focus on the planning part instead of spending time on tedious UI work. A new menu "Planner" with entry "Test Planner" calls into the hard-coded function in planner.c. There a simple dive plan can be constructed with calls to plan_add_segment(&diveplan, duration, depth at the end, fO2, pO2) Calling plan(&diveplan) does the deco calculations and creates deco stops that keep us below the ceiling (with the GFlow/high values currently configured). The stop levels used are defined at the top of planner.c in the stoplevels array - there is no need to do the traditional multiples of 3m or anything like that. The dive including the ascents and deco stops all the way to the surface is completed and then added as simulated dive to the end of the divelist (I guess we could automatically select it later) and can be viewed. This is crude but shows the direction we can go with this. Envision a nice UI that allows you to simply enter the segments and pick the desired stops. What is missing is the ability to give the algorithm additional gases that it can use during the deco phase - right now it simply keeps using the last gas used in the diveplan. All that said, there are clear bugs here - and sadly they seem to be in the deco calculations, as with the example given the ceiling that is calculated makes no sense. When displayed in smooth mode it has very strange jumps up and down that I wouldn't expect. For example with GF 35/75 (the default) the deco ceiling when looking at the simulated dive jumps from 16m back up to 13m around 14:10 into the dive. That seems very odd. Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2013-01-05 07:11:42 +00:00
}
/* overwrite the data in dive
* return false if something goes wrong */
static void create_dive_from_plan(struct diveplan *diveplan, struct dive *dive, bool track_gas)
First stab at simplistic dive planning This comes with absolutely no gui - so the plan literally needs to be compiled into Subsurface. Not exactly a feature, but this allowed me to focus on the planning part instead of spending time on tedious UI work. A new menu "Planner" with entry "Test Planner" calls into the hard-coded function in planner.c. There a simple dive plan can be constructed with calls to plan_add_segment(&diveplan, duration, depth at the end, fO2, pO2) Calling plan(&diveplan) does the deco calculations and creates deco stops that keep us below the ceiling (with the GFlow/high values currently configured). The stop levels used are defined at the top of planner.c in the stoplevels array - there is no need to do the traditional multiples of 3m or anything like that. The dive including the ascents and deco stops all the way to the surface is completed and then added as simulated dive to the end of the divelist (I guess we could automatically select it later) and can be viewed. This is crude but shows the direction we can go with this. Envision a nice UI that allows you to simply enter the segments and pick the desired stops. What is missing is the ability to give the algorithm additional gases that it can use during the deco phase - right now it simply keeps using the last gas used in the diveplan. All that said, there are clear bugs here - and sadly they seem to be in the deco calculations, as with the example given the ceiling that is calculated makes no sense. When displayed in smooth mode it has very strange jumps up and down that I wouldn't expect. For example with GF 35/75 (the default) the deco ceiling when looking at the simulated dive jumps from 16m back up to 13m around 14:10 into the dive. That seems very odd. Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2013-01-05 07:11:42 +00:00
{
struct divedatapoint *dp;
struct divecomputer *dc;
struct sample *sample;
struct event *ev;
cylinder_t *cyl;
int oldpo2 = 0;
Improve profile display in planner This patch allows the planner to save the last manually-entered dive planner point of a dive plan. When the plan has been saved and re-opened for edit, the time of the last-entered dive planner point is used to ensure that dive planning continues from the same point in the profile as was when the original dive plan was saved. Mechanism: 1) In dive.h, create a new dc attribute dc->last_manual_time with data type of duration_t. 2) In diveplanner.c, ensure that the last manually-entered dive planner point is saved in dc->last_manual_time. 3) In save-xml.c, create a new XML attribute for the <divecomputer> element, named last-manual-time. For dive plans, the element would now look like: <divecomputer model='planned dive' last-manual-time='31:17 min'> 4) In parse-xml.c, insert code that recognises the last-manual-time XML attribute, reads the time value and assigns this time to dc->last_manual_time. 5) In diveplannermodel.cpp, method DiveplannerPointModel::loadfromdive, insert code that sets the appropriate boolean value to dp->entered by comparing newtime (i.e. time of dp) with dc->last_manual_time. 6) Diveplannermodel.cpp also accepts profile data from normal dives in the dive log, whether hand-entered or loaded from dive computer. It looks like the reduction of dive points for dives with >100 points continues to work ok. The result is that when a dive plan is saved with manually entered points up to e.g. 10 minutes into the dive, it can be re-opened for edit in the dive planner and the planner re-creates the plan with manually entered points up to 10 minutes. The rest of the points are "soft" points, shaped by the deco calculations of the planner. Improvements: Improve code for profile display in dive planner This responds to #1052. Change load-git.c and save-git.c so that the last-manual-time is also saved in the git-format dive log. Several stylistic changes in text for consistent C source code. Improvement of dive planner profile display: Do some simplification of my alterations to diveplannermodel.cpp Two small style changes in planner.c and diveplannermodel.cpp as requested ny @neolit123 Signed-off-by: Willem Ferguson <willemferguson@zoology.up.ac.za>
2018-01-15 12:51:47 +00:00
int lasttime = 0, last_manual_point = 0;
depth_t lastdepth = {.mm = 0};
int lastcylid;
enum divemode_t type = dive->dc.divemode;
First stab at simplistic dive planning This comes with absolutely no gui - so the plan literally needs to be compiled into Subsurface. Not exactly a feature, but this allowed me to focus on the planning part instead of spending time on tedious UI work. A new menu "Planner" with entry "Test Planner" calls into the hard-coded function in planner.c. There a simple dive plan can be constructed with calls to plan_add_segment(&diveplan, duration, depth at the end, fO2, pO2) Calling plan(&diveplan) does the deco calculations and creates deco stops that keep us below the ceiling (with the GFlow/high values currently configured). The stop levels used are defined at the top of planner.c in the stoplevels array - there is no need to do the traditional multiples of 3m or anything like that. The dive including the ascents and deco stops all the way to the surface is completed and then added as simulated dive to the end of the divelist (I guess we could automatically select it later) and can be viewed. This is crude but shows the direction we can go with this. Envision a nice UI that allows you to simply enter the segments and pick the desired stops. What is missing is the ability to give the algorithm additional gases that it can use during the deco phase - right now it simply keeps using the last gas used in the diveplan. All that said, there are clear bugs here - and sadly they seem to be in the deco calculations, as with the example given the ceiling that is calculated makes no sense. When displayed in smooth mode it has very strange jumps up and down that I wouldn't expect. For example with GF 35/75 (the default) the deco ceiling when looking at the simulated dive jumps from 16m back up to 13m around 14:10 into the dive. That seems very odd. Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2013-01-05 07:11:42 +00:00
if (!diveplan || !diveplan->dp)
return;
#if DEBUG_PLAN & 4
printf("in create_dive_from_plan\n");
dump_plan(diveplan);
#endif
dive->salinity = diveplan->salinity;
// reset the cylinders and clear out the samples and events of the
// dive-to-be-planned so we can restart
reset_cylinders(dive, track_gas);
dc = &dive->dc;
dc->when = dive->when = diveplan->when;
dc->surface_pressure.mbar = diveplan->surface_pressure;
dc->salinity = diveplan->salinity;
free_samples(dc);
while ((ev = dc->events)) {
dc->events = dc->events->next;
free(ev);
}
First stab at simplistic dive planning This comes with absolutely no gui - so the plan literally needs to be compiled into Subsurface. Not exactly a feature, but this allowed me to focus on the planning part instead of spending time on tedious UI work. A new menu "Planner" with entry "Test Planner" calls into the hard-coded function in planner.c. There a simple dive plan can be constructed with calls to plan_add_segment(&diveplan, duration, depth at the end, fO2, pO2) Calling plan(&diveplan) does the deco calculations and creates deco stops that keep us below the ceiling (with the GFlow/high values currently configured). The stop levels used are defined at the top of planner.c in the stoplevels array - there is no need to do the traditional multiples of 3m or anything like that. The dive including the ascents and deco stops all the way to the surface is completed and then added as simulated dive to the end of the divelist (I guess we could automatically select it later) and can be viewed. This is crude but shows the direction we can go with this. Envision a nice UI that allows you to simply enter the segments and pick the desired stops. What is missing is the ability to give the algorithm additional gases that it can use during the deco phase - right now it simply keeps using the last gas used in the diveplan. All that said, there are clear bugs here - and sadly they seem to be in the deco calculations, as with the example given the ceiling that is calculated makes no sense. When displayed in smooth mode it has very strange jumps up and down that I wouldn't expect. For example with GF 35/75 (the default) the deco ceiling when looking at the simulated dive jumps from 16m back up to 13m around 14:10 into the dive. That seems very odd. Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2013-01-05 07:11:42 +00:00
dp = diveplan->dp;
/* Create first sample at time = 0, not based on dp because
* there is no real dp for time = 0, set first cylinder to 0
* O2 setpoint for this sample will be filled later from next dp */
cyl = get_or_create_cylinder(dive, 0);
sample = prepare_sample(dc);
sample->sac.mliter = prefs.bottomsac;
if (track_gas && cyl->type.workingpressure.mbar)
sample->pressure[0].mbar = cyl->end.mbar;
sample->manually_entered = true;
finish_sample(dc);
lastcylid = 0;
while (dp) {
int po2 = dp->setpoint;
int time = dp->time;
depth_t depth = dp->depth;
if (time == 0) {
/* special entries that just inform the algorithm about
* additional gases that are available */
dp = dp->next;
continue;
}
/* Check for SetPoint change */
if (oldpo2 != po2) {
/* this is a bad idea - we should get a different SAMPLE_EVENT type
* reserved for this in libdivecomputer... overloading SMAPLE_EVENT_PO2
* with a different meaning will only cause confusion elsewhere in the code */
if (dp->divemode == type)
add_event(dc, lasttime, SAMPLE_EVENT_PO2, 0, po2, QT_TRANSLATE_NOOP("gettextFromC", "SP change"));
oldpo2 = po2;
}
/* Make sure we have the new gas, and create a gas change event */
if (dp->cylinderid != lastcylid) {
/* need to insert a first sample for the new gas */
add_gas_switch_event(dive, dc, lasttime + 1, dp->cylinderid);
cyl = get_or_create_cylinder(dive, dp->cylinderid);
sample = prepare_sample(dc);
sample[-1].setpoint.mbar = po2;
sample->time.seconds = lasttime + 1;
sample->depth = lastdepth;
sample->manually_entered = dp->entered;
sample->sac.mliter = dp->entered ? prefs.bottomsac : prefs.decosac;
finish_sample(dc);
lastcylid = dp->cylinderid;
First stab at simplistic dive planning This comes with absolutely no gui - so the plan literally needs to be compiled into Subsurface. Not exactly a feature, but this allowed me to focus on the planning part instead of spending time on tedious UI work. A new menu "Planner" with entry "Test Planner" calls into the hard-coded function in planner.c. There a simple dive plan can be constructed with calls to plan_add_segment(&diveplan, duration, depth at the end, fO2, pO2) Calling plan(&diveplan) does the deco calculations and creates deco stops that keep us below the ceiling (with the GFlow/high values currently configured). The stop levels used are defined at the top of planner.c in the stoplevels array - there is no need to do the traditional multiples of 3m or anything like that. The dive including the ascents and deco stops all the way to the surface is completed and then added as simulated dive to the end of the divelist (I guess we could automatically select it later) and can be viewed. This is crude but shows the direction we can go with this. Envision a nice UI that allows you to simply enter the segments and pick the desired stops. What is missing is the ability to give the algorithm additional gases that it can use during the deco phase - right now it simply keeps using the last gas used in the diveplan. All that said, there are clear bugs here - and sadly they seem to be in the deco calculations, as with the example given the ceiling that is calculated makes no sense. When displayed in smooth mode it has very strange jumps up and down that I wouldn't expect. For example with GF 35/75 (the default) the deco ceiling when looking at the simulated dive jumps from 16m back up to 13m around 14:10 into the dive. That seems very odd. Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2013-01-05 07:11:42 +00:00
}
if (dp->divemode != type) {
type = dp->divemode;
add_event(dc, lasttime, 8, 0, type, "modechange");
}
/* Create sample */
First stab at simplistic dive planning This comes with absolutely no gui - so the plan literally needs to be compiled into Subsurface. Not exactly a feature, but this allowed me to focus on the planning part instead of spending time on tedious UI work. A new menu "Planner" with entry "Test Planner" calls into the hard-coded function in planner.c. There a simple dive plan can be constructed with calls to plan_add_segment(&diveplan, duration, depth at the end, fO2, pO2) Calling plan(&diveplan) does the deco calculations and creates deco stops that keep us below the ceiling (with the GFlow/high values currently configured). The stop levels used are defined at the top of planner.c in the stoplevels array - there is no need to do the traditional multiples of 3m or anything like that. The dive including the ascents and deco stops all the way to the surface is completed and then added as simulated dive to the end of the divelist (I guess we could automatically select it later) and can be viewed. This is crude but shows the direction we can go with this. Envision a nice UI that allows you to simply enter the segments and pick the desired stops. What is missing is the ability to give the algorithm additional gases that it can use during the deco phase - right now it simply keeps using the last gas used in the diveplan. All that said, there are clear bugs here - and sadly they seem to be in the deco calculations, as with the example given the ceiling that is calculated makes no sense. When displayed in smooth mode it has very strange jumps up and down that I wouldn't expect. For example with GF 35/75 (the default) the deco ceiling when looking at the simulated dive jumps from 16m back up to 13m around 14:10 into the dive. That seems very odd. Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2013-01-05 07:11:42 +00:00
sample = prepare_sample(dc);
/* set po2 at beginning of this segment */
/* and keep it valid for last sample - where it likely doesn't matter */
sample[-1].setpoint.mbar = po2;
sample->setpoint.mbar = po2;
sample->time.seconds = lasttime = time;
Improve profile display in planner This patch allows the planner to save the last manually-entered dive planner point of a dive plan. When the plan has been saved and re-opened for edit, the time of the last-entered dive planner point is used to ensure that dive planning continues from the same point in the profile as was when the original dive plan was saved. Mechanism: 1) In dive.h, create a new dc attribute dc->last_manual_time with data type of duration_t. 2) In diveplanner.c, ensure that the last manually-entered dive planner point is saved in dc->last_manual_time. 3) In save-xml.c, create a new XML attribute for the <divecomputer> element, named last-manual-time. For dive plans, the element would now look like: <divecomputer model='planned dive' last-manual-time='31:17 min'> 4) In parse-xml.c, insert code that recognises the last-manual-time XML attribute, reads the time value and assigns this time to dc->last_manual_time. 5) In diveplannermodel.cpp, method DiveplannerPointModel::loadfromdive, insert code that sets the appropriate boolean value to dp->entered by comparing newtime (i.e. time of dp) with dc->last_manual_time. 6) Diveplannermodel.cpp also accepts profile data from normal dives in the dive log, whether hand-entered or loaded from dive computer. It looks like the reduction of dive points for dives with >100 points continues to work ok. The result is that when a dive plan is saved with manually entered points up to e.g. 10 minutes into the dive, it can be re-opened for edit in the dive planner and the planner re-creates the plan with manually entered points up to 10 minutes. The rest of the points are "soft" points, shaped by the deco calculations of the planner. Improvements: Improve code for profile display in dive planner This responds to #1052. Change load-git.c and save-git.c so that the last-manual-time is also saved in the git-format dive log. Several stylistic changes in text for consistent C source code. Improvement of dive planner profile display: Do some simplification of my alterations to diveplannermodel.cpp Two small style changes in planner.c and diveplannermodel.cpp as requested ny @neolit123 Signed-off-by: Willem Ferguson <willemferguson@zoology.up.ac.za>
2018-01-15 12:51:47 +00:00
if (dp->entered) last_manual_point = dp->time;
sample->depth = lastdepth = depth;
sample->manually_entered = dp->entered;
sample->sac.mliter = dp->entered ? prefs.bottomsac : prefs.decosac;
if (track_gas && !sample[-1].setpoint.mbar) { /* Don't track gas usage for CCR legs of dive */
update_cylinder_pressure(dive, sample[-1].depth.mm, depth.mm, time - sample[-1].time.seconds,
dp->entered ? diveplan->bottomsac : diveplan->decosac, cyl, !dp->entered, type);
if (cyl->type.workingpressure.mbar)
sample->pressure[0].mbar = cyl->end.mbar;
}
finish_sample(dc);
First stab at simplistic dive planning This comes with absolutely no gui - so the plan literally needs to be compiled into Subsurface. Not exactly a feature, but this allowed me to focus on the planning part instead of spending time on tedious UI work. A new menu "Planner" with entry "Test Planner" calls into the hard-coded function in planner.c. There a simple dive plan can be constructed with calls to plan_add_segment(&diveplan, duration, depth at the end, fO2, pO2) Calling plan(&diveplan) does the deco calculations and creates deco stops that keep us below the ceiling (with the GFlow/high values currently configured). The stop levels used are defined at the top of planner.c in the stoplevels array - there is no need to do the traditional multiples of 3m or anything like that. The dive including the ascents and deco stops all the way to the surface is completed and then added as simulated dive to the end of the divelist (I guess we could automatically select it later) and can be viewed. This is crude but shows the direction we can go with this. Envision a nice UI that allows you to simply enter the segments and pick the desired stops. What is missing is the ability to give the algorithm additional gases that it can use during the deco phase - right now it simply keeps using the last gas used in the diveplan. All that said, there are clear bugs here - and sadly they seem to be in the deco calculations, as with the example given the ceiling that is calculated makes no sense. When displayed in smooth mode it has very strange jumps up and down that I wouldn't expect. For example with GF 35/75 (the default) the deco ceiling when looking at the simulated dive jumps from 16m back up to 13m around 14:10 into the dive. That seems very odd. Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2013-01-05 07:11:42 +00:00
dp = dp->next;
}
Improve profile display in planner This patch allows the planner to save the last manually-entered dive planner point of a dive plan. When the plan has been saved and re-opened for edit, the time of the last-entered dive planner point is used to ensure that dive planning continues from the same point in the profile as was when the original dive plan was saved. Mechanism: 1) In dive.h, create a new dc attribute dc->last_manual_time with data type of duration_t. 2) In diveplanner.c, ensure that the last manually-entered dive planner point is saved in dc->last_manual_time. 3) In save-xml.c, create a new XML attribute for the <divecomputer> element, named last-manual-time. For dive plans, the element would now look like: <divecomputer model='planned dive' last-manual-time='31:17 min'> 4) In parse-xml.c, insert code that recognises the last-manual-time XML attribute, reads the time value and assigns this time to dc->last_manual_time. 5) In diveplannermodel.cpp, method DiveplannerPointModel::loadfromdive, insert code that sets the appropriate boolean value to dp->entered by comparing newtime (i.e. time of dp) with dc->last_manual_time. 6) Diveplannermodel.cpp also accepts profile data from normal dives in the dive log, whether hand-entered or loaded from dive computer. It looks like the reduction of dive points for dives with >100 points continues to work ok. The result is that when a dive plan is saved with manually entered points up to e.g. 10 minutes into the dive, it can be re-opened for edit in the dive planner and the planner re-creates the plan with manually entered points up to 10 minutes. The rest of the points are "soft" points, shaped by the deco calculations of the planner. Improvements: Improve code for profile display in dive planner This responds to #1052. Change load-git.c and save-git.c so that the last-manual-time is also saved in the git-format dive log. Several stylistic changes in text for consistent C source code. Improvement of dive planner profile display: Do some simplification of my alterations to diveplannermodel.cpp Two small style changes in planner.c and diveplannermodel.cpp as requested ny @neolit123 Signed-off-by: Willem Ferguson <willemferguson@zoology.up.ac.za>
2018-01-15 12:51:47 +00:00
dive->dc.last_manual_time.seconds = last_manual_point;
#if DEBUG_PLAN & 32
save_dive(stdout, dive);
#endif
return;
First stab at simplistic dive planning This comes with absolutely no gui - so the plan literally needs to be compiled into Subsurface. Not exactly a feature, but this allowed me to focus on the planning part instead of spending time on tedious UI work. A new menu "Planner" with entry "Test Planner" calls into the hard-coded function in planner.c. There a simple dive plan can be constructed with calls to plan_add_segment(&diveplan, duration, depth at the end, fO2, pO2) Calling plan(&diveplan) does the deco calculations and creates deco stops that keep us below the ceiling (with the GFlow/high values currently configured). The stop levels used are defined at the top of planner.c in the stoplevels array - there is no need to do the traditional multiples of 3m or anything like that. The dive including the ascents and deco stops all the way to the surface is completed and then added as simulated dive to the end of the divelist (I guess we could automatically select it later) and can be viewed. This is crude but shows the direction we can go with this. Envision a nice UI that allows you to simply enter the segments and pick the desired stops. What is missing is the ability to give the algorithm additional gases that it can use during the deco phase - right now it simply keeps using the last gas used in the diveplan. All that said, there are clear bugs here - and sadly they seem to be in the deco calculations, as with the example given the ceiling that is calculated makes no sense. When displayed in smooth mode it has very strange jumps up and down that I wouldn't expect. For example with GF 35/75 (the default) the deco ceiling when looking at the simulated dive jumps from 16m back up to 13m around 14:10 into the dive. That seems very odd. Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2013-01-05 07:11:42 +00:00
}
void free_dps(struct diveplan *diveplan)
{
if (!diveplan)
return;
struct divedatapoint *dp = diveplan->dp;
while (dp) {
struct divedatapoint *ndp = dp->next;
free(dp);
dp = ndp;
}
diveplan->dp = NULL;
}
struct divedatapoint *create_dp(int time_incr, int depth, int cylinderid, int po2)
First stab at simplistic dive planning This comes with absolutely no gui - so the plan literally needs to be compiled into Subsurface. Not exactly a feature, but this allowed me to focus on the planning part instead of spending time on tedious UI work. A new menu "Planner" with entry "Test Planner" calls into the hard-coded function in planner.c. There a simple dive plan can be constructed with calls to plan_add_segment(&diveplan, duration, depth at the end, fO2, pO2) Calling plan(&diveplan) does the deco calculations and creates deco stops that keep us below the ceiling (with the GFlow/high values currently configured). The stop levels used are defined at the top of planner.c in the stoplevels array - there is no need to do the traditional multiples of 3m or anything like that. The dive including the ascents and deco stops all the way to the surface is completed and then added as simulated dive to the end of the divelist (I guess we could automatically select it later) and can be viewed. This is crude but shows the direction we can go with this. Envision a nice UI that allows you to simply enter the segments and pick the desired stops. What is missing is the ability to give the algorithm additional gases that it can use during the deco phase - right now it simply keeps using the last gas used in the diveplan. All that said, there are clear bugs here - and sadly they seem to be in the deco calculations, as with the example given the ceiling that is calculated makes no sense. When displayed in smooth mode it has very strange jumps up and down that I wouldn't expect. For example with GF 35/75 (the default) the deco ceiling when looking at the simulated dive jumps from 16m back up to 13m around 14:10 into the dive. That seems very odd. Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2013-01-05 07:11:42 +00:00
{
struct divedatapoint *dp;
dp = malloc(sizeof(struct divedatapoint));
dp->time = time_incr;
dp->depth.mm = depth;
dp->cylinderid = cylinderid;
dp->minimum_gas.mbar = 0;
dp->setpoint = po2;
dp->entered = false;
First stab at simplistic dive planning This comes with absolutely no gui - so the plan literally needs to be compiled into Subsurface. Not exactly a feature, but this allowed me to focus on the planning part instead of spending time on tedious UI work. A new menu "Planner" with entry "Test Planner" calls into the hard-coded function in planner.c. There a simple dive plan can be constructed with calls to plan_add_segment(&diveplan, duration, depth at the end, fO2, pO2) Calling plan(&diveplan) does the deco calculations and creates deco stops that keep us below the ceiling (with the GFlow/high values currently configured). The stop levels used are defined at the top of planner.c in the stoplevels array - there is no need to do the traditional multiples of 3m or anything like that. The dive including the ascents and deco stops all the way to the surface is completed and then added as simulated dive to the end of the divelist (I guess we could automatically select it later) and can be viewed. This is crude but shows the direction we can go with this. Envision a nice UI that allows you to simply enter the segments and pick the desired stops. What is missing is the ability to give the algorithm additional gases that it can use during the deco phase - right now it simply keeps using the last gas used in the diveplan. All that said, there are clear bugs here - and sadly they seem to be in the deco calculations, as with the example given the ceiling that is calculated makes no sense. When displayed in smooth mode it has very strange jumps up and down that I wouldn't expect. For example with GF 35/75 (the default) the deco ceiling when looking at the simulated dive jumps from 16m back up to 13m around 14:10 into the dive. That seems very odd. Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2013-01-05 07:11:42 +00:00
dp->next = NULL;
return dp;
}
static void add_to_end_of_diveplan(struct diveplan *diveplan, struct divedatapoint *dp)
First stab at simplistic dive planning This comes with absolutely no gui - so the plan literally needs to be compiled into Subsurface. Not exactly a feature, but this allowed me to focus on the planning part instead of spending time on tedious UI work. A new menu "Planner" with entry "Test Planner" calls into the hard-coded function in planner.c. There a simple dive plan can be constructed with calls to plan_add_segment(&diveplan, duration, depth at the end, fO2, pO2) Calling plan(&diveplan) does the deco calculations and creates deco stops that keep us below the ceiling (with the GFlow/high values currently configured). The stop levels used are defined at the top of planner.c in the stoplevels array - there is no need to do the traditional multiples of 3m or anything like that. The dive including the ascents and deco stops all the way to the surface is completed and then added as simulated dive to the end of the divelist (I guess we could automatically select it later) and can be viewed. This is crude but shows the direction we can go with this. Envision a nice UI that allows you to simply enter the segments and pick the desired stops. What is missing is the ability to give the algorithm additional gases that it can use during the deco phase - right now it simply keeps using the last gas used in the diveplan. All that said, there are clear bugs here - and sadly they seem to be in the deco calculations, as with the example given the ceiling that is calculated makes no sense. When displayed in smooth mode it has very strange jumps up and down that I wouldn't expect. For example with GF 35/75 (the default) the deco ceiling when looking at the simulated dive jumps from 16m back up to 13m around 14:10 into the dive. That seems very odd. Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2013-01-05 07:11:42 +00:00
{
struct divedatapoint **lastdp = &diveplan->dp;
struct divedatapoint *ldp = *lastdp;
int lasttime = 0;
while (*lastdp) {
First stab at simplistic dive planning This comes with absolutely no gui - so the plan literally needs to be compiled into Subsurface. Not exactly a feature, but this allowed me to focus on the planning part instead of spending time on tedious UI work. A new menu "Planner" with entry "Test Planner" calls into the hard-coded function in planner.c. There a simple dive plan can be constructed with calls to plan_add_segment(&diveplan, duration, depth at the end, fO2, pO2) Calling plan(&diveplan) does the deco calculations and creates deco stops that keep us below the ceiling (with the GFlow/high values currently configured). The stop levels used are defined at the top of planner.c in the stoplevels array - there is no need to do the traditional multiples of 3m or anything like that. The dive including the ascents and deco stops all the way to the surface is completed and then added as simulated dive to the end of the divelist (I guess we could automatically select it later) and can be viewed. This is crude but shows the direction we can go with this. Envision a nice UI that allows you to simply enter the segments and pick the desired stops. What is missing is the ability to give the algorithm additional gases that it can use during the deco phase - right now it simply keeps using the last gas used in the diveplan. All that said, there are clear bugs here - and sadly they seem to be in the deco calculations, as with the example given the ceiling that is calculated makes no sense. When displayed in smooth mode it has very strange jumps up and down that I wouldn't expect. For example with GF 35/75 (the default) the deco ceiling when looking at the simulated dive jumps from 16m back up to 13m around 14:10 into the dive. That seems very odd. Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2013-01-05 07:11:42 +00:00
ldp = *lastdp;
if (ldp->time > lasttime)
lasttime = ldp->time;
First stab at simplistic dive planning This comes with absolutely no gui - so the plan literally needs to be compiled into Subsurface. Not exactly a feature, but this allowed me to focus on the planning part instead of spending time on tedious UI work. A new menu "Planner" with entry "Test Planner" calls into the hard-coded function in planner.c. There a simple dive plan can be constructed with calls to plan_add_segment(&diveplan, duration, depth at the end, fO2, pO2) Calling plan(&diveplan) does the deco calculations and creates deco stops that keep us below the ceiling (with the GFlow/high values currently configured). The stop levels used are defined at the top of planner.c in the stoplevels array - there is no need to do the traditional multiples of 3m or anything like that. The dive including the ascents and deco stops all the way to the surface is completed and then added as simulated dive to the end of the divelist (I guess we could automatically select it later) and can be viewed. This is crude but shows the direction we can go with this. Envision a nice UI that allows you to simply enter the segments and pick the desired stops. What is missing is the ability to give the algorithm additional gases that it can use during the deco phase - right now it simply keeps using the last gas used in the diveplan. All that said, there are clear bugs here - and sadly they seem to be in the deco calculations, as with the example given the ceiling that is calculated makes no sense. When displayed in smooth mode it has very strange jumps up and down that I wouldn't expect. For example with GF 35/75 (the default) the deco ceiling when looking at the simulated dive jumps from 16m back up to 13m around 14:10 into the dive. That seems very odd. Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2013-01-05 07:11:42 +00:00
lastdp = &(*lastdp)->next;
}
*lastdp = dp;
if (ldp)
dp->time += lasttime;
First stab at simplistic dive planning This comes with absolutely no gui - so the plan literally needs to be compiled into Subsurface. Not exactly a feature, but this allowed me to focus on the planning part instead of spending time on tedious UI work. A new menu "Planner" with entry "Test Planner" calls into the hard-coded function in planner.c. There a simple dive plan can be constructed with calls to plan_add_segment(&diveplan, duration, depth at the end, fO2, pO2) Calling plan(&diveplan) does the deco calculations and creates deco stops that keep us below the ceiling (with the GFlow/high values currently configured). The stop levels used are defined at the top of planner.c in the stoplevels array - there is no need to do the traditional multiples of 3m or anything like that. The dive including the ascents and deco stops all the way to the surface is completed and then added as simulated dive to the end of the divelist (I guess we could automatically select it later) and can be viewed. This is crude but shows the direction we can go with this. Envision a nice UI that allows you to simply enter the segments and pick the desired stops. What is missing is the ability to give the algorithm additional gases that it can use during the deco phase - right now it simply keeps using the last gas used in the diveplan. All that said, there are clear bugs here - and sadly they seem to be in the deco calculations, as with the example given the ceiling that is calculated makes no sense. When displayed in smooth mode it has very strange jumps up and down that I wouldn't expect. For example with GF 35/75 (the default) the deco ceiling when looking at the simulated dive jumps from 16m back up to 13m around 14:10 into the dive. That seems very odd. Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2013-01-05 07:11:42 +00:00
}
struct divedatapoint *plan_add_segment(struct diveplan *diveplan, int duration, int depth, int cylinderid, int po2, bool entered, enum divemode_t divemode)
First stab at simplistic dive planning This comes with absolutely no gui - so the plan literally needs to be compiled into Subsurface. Not exactly a feature, but this allowed me to focus on the planning part instead of spending time on tedious UI work. A new menu "Planner" with entry "Test Planner" calls into the hard-coded function in planner.c. There a simple dive plan can be constructed with calls to plan_add_segment(&diveplan, duration, depth at the end, fO2, pO2) Calling plan(&diveplan) does the deco calculations and creates deco stops that keep us below the ceiling (with the GFlow/high values currently configured). The stop levels used are defined at the top of planner.c in the stoplevels array - there is no need to do the traditional multiples of 3m or anything like that. The dive including the ascents and deco stops all the way to the surface is completed and then added as simulated dive to the end of the divelist (I guess we could automatically select it later) and can be viewed. This is crude but shows the direction we can go with this. Envision a nice UI that allows you to simply enter the segments and pick the desired stops. What is missing is the ability to give the algorithm additional gases that it can use during the deco phase - right now it simply keeps using the last gas used in the diveplan. All that said, there are clear bugs here - and sadly they seem to be in the deco calculations, as with the example given the ceiling that is calculated makes no sense. When displayed in smooth mode it has very strange jumps up and down that I wouldn't expect. For example with GF 35/75 (the default) the deco ceiling when looking at the simulated dive jumps from 16m back up to 13m around 14:10 into the dive. That seems very odd. Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2013-01-05 07:11:42 +00:00
{
struct divedatapoint *dp = create_dp(duration, depth, cylinderid, divemode == CCR ? po2 : 0);
dp->entered = entered;
dp->divemode = divemode;
First stab at simplistic dive planning This comes with absolutely no gui - so the plan literally needs to be compiled into Subsurface. Not exactly a feature, but this allowed me to focus on the planning part instead of spending time on tedious UI work. A new menu "Planner" with entry "Test Planner" calls into the hard-coded function in planner.c. There a simple dive plan can be constructed with calls to plan_add_segment(&diveplan, duration, depth at the end, fO2, pO2) Calling plan(&diveplan) does the deco calculations and creates deco stops that keep us below the ceiling (with the GFlow/high values currently configured). The stop levels used are defined at the top of planner.c in the stoplevels array - there is no need to do the traditional multiples of 3m or anything like that. The dive including the ascents and deco stops all the way to the surface is completed and then added as simulated dive to the end of the divelist (I guess we could automatically select it later) and can be viewed. This is crude but shows the direction we can go with this. Envision a nice UI that allows you to simply enter the segments and pick the desired stops. What is missing is the ability to give the algorithm additional gases that it can use during the deco phase - right now it simply keeps using the last gas used in the diveplan. All that said, there are clear bugs here - and sadly they seem to be in the deco calculations, as with the example given the ceiling that is calculated makes no sense. When displayed in smooth mode it has very strange jumps up and down that I wouldn't expect. For example with GF 35/75 (the default) the deco ceiling when looking at the simulated dive jumps from 16m back up to 13m around 14:10 into the dive. That seems very odd. Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2013-01-05 07:11:42 +00:00
add_to_end_of_diveplan(diveplan, dp);
return dp;
First stab at simplistic dive planning This comes with absolutely no gui - so the plan literally needs to be compiled into Subsurface. Not exactly a feature, but this allowed me to focus on the planning part instead of spending time on tedious UI work. A new menu "Planner" with entry "Test Planner" calls into the hard-coded function in planner.c. There a simple dive plan can be constructed with calls to plan_add_segment(&diveplan, duration, depth at the end, fO2, pO2) Calling plan(&diveplan) does the deco calculations and creates deco stops that keep us below the ceiling (with the GFlow/high values currently configured). The stop levels used are defined at the top of planner.c in the stoplevels array - there is no need to do the traditional multiples of 3m or anything like that. The dive including the ascents and deco stops all the way to the surface is completed and then added as simulated dive to the end of the divelist (I guess we could automatically select it later) and can be viewed. This is crude but shows the direction we can go with this. Envision a nice UI that allows you to simply enter the segments and pick the desired stops. What is missing is the ability to give the algorithm additional gases that it can use during the deco phase - right now it simply keeps using the last gas used in the diveplan. All that said, there are clear bugs here - and sadly they seem to be in the deco calculations, as with the example given the ceiling that is calculated makes no sense. When displayed in smooth mode it has very strange jumps up and down that I wouldn't expect. For example with GF 35/75 (the default) the deco ceiling when looking at the simulated dive jumps from 16m back up to 13m around 14:10 into the dive. That seems very odd. Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2013-01-05 07:11:42 +00:00
}
struct gaschanges {
int depth;
int gasidx;
};
static struct gaschanges *analyze_gaslist(struct diveplan *diveplan, struct dive *dive, int *gaschangenr, int depth, int *asc_cylinder)
{
int nr = 0;
struct gaschanges *gaschanges = NULL;
struct divedatapoint *dp = diveplan->dp;
int best_depth = get_cylinder(dive, *asc_cylinder)->depth.mm;
bool total_time_zero = true;
while (dp) {
if (dp->time == 0 && total_time_zero) {
if (dp->depth.mm <= depth) {
int i = 0;
nr++;
gaschanges = realloc(gaschanges, nr * sizeof(struct gaschanges));
while (i < nr - 1) {
if (dp->depth.mm < gaschanges[i].depth) {
memmove(gaschanges + i + 1, gaschanges + i, (nr - i - 1) * sizeof(struct gaschanges));
break;
}
i++;
}
gaschanges[i].depth = dp->depth.mm;
gaschanges[i].gasidx = dp->cylinderid;
assert(gaschanges[i].gasidx != -1);
} else {
/* is there a better mix to start deco? */
if (dp->depth.mm < best_depth) {
best_depth = dp->depth.mm;
*asc_cylinder = dp->cylinderid;
}
}
} else {
total_time_zero = false;
}
dp = dp->next;
}
*gaschangenr = nr;
#if DEBUG_PLAN & 16
for (nr = 0; nr < *gaschangenr; nr++) {
int idx = gaschanges[nr].gasidx;
printf("gaschange nr %d: @ %5.2lfm gasidx %d (%s)\n", nr, gaschanges[nr].depth / 1000.0,
idx, gasname(&get_cylinder(&dive, idx)->gasmix));
}
#endif
return gaschanges;
}
/* sort all the stops into one ordered list */
static int *sort_stops(int *dstops, int dnr, struct gaschanges *gstops, int gnr)
{
int i, gi, di;
int total = dnr + gnr;
int *stoplevels = malloc(total * sizeof(int));
/* no gaschanges */
if (gnr == 0) {
Improvement for various bits of the planner Rewrite of the actual planner logic. Now ascend to the next potential stop depth. There the state is cached and we try to ascend to the next stop depth. If we hit the ceiling while doing that we go back to the cached state and wait there for a minute. Then we try again. Then loop. Converted all depth related variables from unsigned int to int. During planning, in a time step the current depth can temporarily be negative and comparisons of a negative int with an unsigned it have not the result I expected ( (int) -2 < (unsigned int) 3 turns out to be false). And we don’t really need the 32nd bit that unsigned buys us for depths. Deco stops are now shown in the same table as manually entered stops in boldface (I removed the second table to save screen estate). The gas shown in the table is still misleading as it means the gas used on the segment leading up to that event. The update of the profile only works partially upon changes in the list of available gases. Treatment of various gases is basically there but needs some more love. The ascent velocity is now provided by a function that takes the current depth as argument. Currently it always returns 10m/min but that will later be variable (and hopefully user configurable). The profile is not redrawn while deco is computed (avoiding an infinite recursion). The table got a new column for the duration of a segment while the old “duration” column was renamed “Runtime” to reflect what it actually shows. Currently, only the run time but not the duration are editable. All deco gases are used from the depth where their pO2 is 1.4bar. This should become more flexible. Calculation of the pressure drop in cylinders without configured volumes is suppressed. This solves a problem with the planner crashing when saving a dive where not all cylinders had been manually given a volume. [Short rant break: Treating 0/0 as air bites back at so many places. E.g. Cylinder data is initialized with memsetting the whole structures to 0. Then later suddenly this totally unconfigured cylinder is being treated as it would contain air. Maybe at some point this was a feature. But it lead to a naughty bug which took me over an hour to resolve. We should seriously reconsider this choice and better move to 209/0 being air if changing this everywhere is not too much trouble] Signed-off-by: Robert C. Helling <helling@atdotde.de> Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2014-04-17 08:54:55 +00:00
memcpy(stoplevels, dstops, dnr * sizeof(int));
return stoplevels;
}
i = total - 1;
gi = gnr - 1;
di = dnr - 1;
while (i >= 0) {
if (dstops[di] > gstops[gi].depth) {
stoplevels[i] = dstops[di];
di--;
} else if (dstops[di] == gstops[gi].depth) {
stoplevels[i] = dstops[di];
di--;
gi--;
} else {
stoplevels[i] = gstops[gi].depth;
gi--;
}
i--;
if (di < 0) {
while (gi >= 0)
stoplevels[i--] = gstops[gi--].depth;
break;
}
if (gi < 0) {
while (di >= 0)
stoplevels[i--] = dstops[di--];
break;
}
}
while (i >= 0)
stoplevels[i--] = 0;
#if DEBUG_PLAN & 16
int k;
for (k = gnr + dnr - 1; k >= 0; k--) {
printf("stoplevel[%d]: %5.2lfm\n", k, stoplevels[k] / 1000.0);
if (stoplevels[k] == 0)
break;
}
#endif
return stoplevels;
}
int ascent_velocity(int depth, int avg_depth, int bottom_time)
Improvement for various bits of the planner Rewrite of the actual planner logic. Now ascend to the next potential stop depth. There the state is cached and we try to ascend to the next stop depth. If we hit the ceiling while doing that we go back to the cached state and wait there for a minute. Then we try again. Then loop. Converted all depth related variables from unsigned int to int. During planning, in a time step the current depth can temporarily be negative and comparisons of a negative int with an unsigned it have not the result I expected ( (int) -2 < (unsigned int) 3 turns out to be false). And we don’t really need the 32nd bit that unsigned buys us for depths. Deco stops are now shown in the same table as manually entered stops in boldface (I removed the second table to save screen estate). The gas shown in the table is still misleading as it means the gas used on the segment leading up to that event. The update of the profile only works partially upon changes in the list of available gases. Treatment of various gases is basically there but needs some more love. The ascent velocity is now provided by a function that takes the current depth as argument. Currently it always returns 10m/min but that will later be variable (and hopefully user configurable). The profile is not redrawn while deco is computed (avoiding an infinite recursion). The table got a new column for the duration of a segment while the old “duration” column was renamed “Runtime” to reflect what it actually shows. Currently, only the run time but not the duration are editable. All deco gases are used from the depth where their pO2 is 1.4bar. This should become more flexible. Calculation of the pressure drop in cylinders without configured volumes is suppressed. This solves a problem with the planner crashing when saving a dive where not all cylinders had been manually given a volume. [Short rant break: Treating 0/0 as air bites back at so many places. E.g. Cylinder data is initialized with memsetting the whole structures to 0. Then later suddenly this totally unconfigured cylinder is being treated as it would contain air. Maybe at some point this was a feature. But it lead to a naughty bug which took me over an hour to resolve. We should seriously reconsider this choice and better move to 209/0 being air if changing this everywhere is not too much trouble] Signed-off-by: Robert C. Helling <helling@atdotde.de> Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2014-04-17 08:54:55 +00:00
{
UNUSED(bottom_time);
Improvement for various bits of the planner Rewrite of the actual planner logic. Now ascend to the next potential stop depth. There the state is cached and we try to ascend to the next stop depth. If we hit the ceiling while doing that we go back to the cached state and wait there for a minute. Then we try again. Then loop. Converted all depth related variables from unsigned int to int. During planning, in a time step the current depth can temporarily be negative and comparisons of a negative int with an unsigned it have not the result I expected ( (int) -2 < (unsigned int) 3 turns out to be false). And we don’t really need the 32nd bit that unsigned buys us for depths. Deco stops are now shown in the same table as manually entered stops in boldface (I removed the second table to save screen estate). The gas shown in the table is still misleading as it means the gas used on the segment leading up to that event. The update of the profile only works partially upon changes in the list of available gases. Treatment of various gases is basically there but needs some more love. The ascent velocity is now provided by a function that takes the current depth as argument. Currently it always returns 10m/min but that will later be variable (and hopefully user configurable). The profile is not redrawn while deco is computed (avoiding an infinite recursion). The table got a new column for the duration of a segment while the old “duration” column was renamed “Runtime” to reflect what it actually shows. Currently, only the run time but not the duration are editable. All deco gases are used from the depth where their pO2 is 1.4bar. This should become more flexible. Calculation of the pressure drop in cylinders without configured volumes is suppressed. This solves a problem with the planner crashing when saving a dive where not all cylinders had been manually given a volume. [Short rant break: Treating 0/0 as air bites back at so many places. E.g. Cylinder data is initialized with memsetting the whole structures to 0. Then later suddenly this totally unconfigured cylinder is being treated as it would contain air. Maybe at some point this was a feature. But it lead to a naughty bug which took me over an hour to resolve. We should seriously reconsider this choice and better move to 209/0 being air if changing this everywhere is not too much trouble] Signed-off-by: Robert C. Helling <helling@atdotde.de> Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2014-04-17 08:54:55 +00:00
/* We need to make this configurable */
/* As an example (and possibly reasonable default) this is the Tech 1 provedure according
* to http://www.globalunderwaterexplorers.org/files/Standards_and_Procedures/SOP_Manual_Ver2.0.2.pdf */
if (depth * 4 > avg_depth * 3) {
return prefs.ascrate75;
} else {
if (depth * 2 > avg_depth) {
return prefs.ascrate50;
} else {
if (depth > 6000)
return prefs.ascratestops;
else
return prefs.ascratelast6m;
}
}
Improvement for various bits of the planner Rewrite of the actual planner logic. Now ascend to the next potential stop depth. There the state is cached and we try to ascend to the next stop depth. If we hit the ceiling while doing that we go back to the cached state and wait there for a minute. Then we try again. Then loop. Converted all depth related variables from unsigned int to int. During planning, in a time step the current depth can temporarily be negative and comparisons of a negative int with an unsigned it have not the result I expected ( (int) -2 < (unsigned int) 3 turns out to be false). And we don’t really need the 32nd bit that unsigned buys us for depths. Deco stops are now shown in the same table as manually entered stops in boldface (I removed the second table to save screen estate). The gas shown in the table is still misleading as it means the gas used on the segment leading up to that event. The update of the profile only works partially upon changes in the list of available gases. Treatment of various gases is basically there but needs some more love. The ascent velocity is now provided by a function that takes the current depth as argument. Currently it always returns 10m/min but that will later be variable (and hopefully user configurable). The profile is not redrawn while deco is computed (avoiding an infinite recursion). The table got a new column for the duration of a segment while the old “duration” column was renamed “Runtime” to reflect what it actually shows. Currently, only the run time but not the duration are editable. All deco gases are used from the depth where their pO2 is 1.4bar. This should become more flexible. Calculation of the pressure drop in cylinders without configured volumes is suppressed. This solves a problem with the planner crashing when saving a dive where not all cylinders had been manually given a volume. [Short rant break: Treating 0/0 as air bites back at so many places. E.g. Cylinder data is initialized with memsetting the whole structures to 0. Then later suddenly this totally unconfigured cylinder is being treated as it would contain air. Maybe at some point this was a feature. But it lead to a naughty bug which took me over an hour to resolve. We should seriously reconsider this choice and better move to 209/0 being air if changing this everywhere is not too much trouble] Signed-off-by: Robert C. Helling <helling@atdotde.de> Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2014-04-17 08:54:55 +00:00
}
static void track_ascent_gas(int depth, struct dive *dive, int cylinder_id, int avg_depth, int bottom_time, bool safety_stop, enum divemode_t divemode)
{
cylinder_t *cylinder = get_cylinder(dive, cylinder_id);
while (depth > 0) {
int deltad = ascent_velocity(depth, avg_depth, bottom_time) * TIMESTEP;
if (deltad > depth)
deltad = depth;
update_cylinder_pressure(dive, depth, depth - deltad, TIMESTEP, prefs.decosac, cylinder, true, divemode);
if (depth <= 5000 && depth >= (5000 - deltad) && safety_stop) {
update_cylinder_pressure(dive, 5000, 5000, 180, prefs.decosac, cylinder, true, divemode);
safety_stop = false;
}
depth -= deltad;
}
}
// Determine whether ascending to the next stop will break the ceiling. Return true if the ascent is ok, false if it isn't.
static bool trial_ascent(struct deco_state *ds, int wait_time, int trial_depth, int stoplevel, int avg_depth, int bottom_time, struct gasmix gasmix, int po2, double surface_pressure, struct dive *dive, enum divemode_t divemode)
{
bool clear_to_ascend = true;
struct deco_state *trial_cache = NULL;
// For consistency with other VPM-B implementations, we should not start the ascent while the ceiling is
// deeper than the next stop (thus the offgasing during the ascent is ignored).
// However, we still need to make sure we don't break the ceiling due to on-gassing during ascent.
cache_deco_state(ds, &trial_cache);
if (wait_time)
add_segment(ds, depth_to_bar(trial_depth, dive),
gasmix,
wait_time, po2, divemode, prefs.decosac);
if (decoMode() == VPMB) {
double tolerance_limit = tissue_tolerance_calc(ds, dive, depth_to_bar(stoplevel, dive));
update_regression(ds, dive);
if (deco_allowed_depth(tolerance_limit, surface_pressure, dive, 1) > stoplevel) {
restore_deco_state(trial_cache, ds, false);
free(trial_cache);
return false;
}
}
while (trial_depth > stoplevel) {
double tolerance_limit;
int deltad = ascent_velocity(trial_depth, avg_depth, bottom_time) * TIMESTEP;
if (deltad > trial_depth) /* don't test against depth above surface */
deltad = trial_depth;
add_segment(ds, depth_to_bar(trial_depth, dive),
gasmix,
TIMESTEP, po2, divemode, prefs.decosac);
tolerance_limit = tissue_tolerance_calc(ds, dive, depth_to_bar(trial_depth, dive));
if (decoMode() == VPMB)
update_regression(ds, dive);
if (deco_allowed_depth(tolerance_limit, surface_pressure, dive, 1) > trial_depth - deltad) {
/* We should have stopped */
clear_to_ascend = false;
break;
}
trial_depth -= deltad;
}
restore_deco_state(trial_cache, ds, false);
free(trial_cache);
return clear_to_ascend;
}
/* Determine if there is enough gas for the dive. Return true if there is enough.
* Also return true if this cannot be calculated because the cylinder doesn't have
* size or a starting pressure.
*/
static bool enough_gas(const struct dive *dive, int current_cylinder)
{
cylinder_t *cyl;
if (current_cylinder < 0 || current_cylinder >= dive->cylinders.nr)
return false;
cyl = get_cylinder(dive, current_cylinder);
if (!cyl->start.mbar)
return true;
if (cyl->type.size.mliter)
Unify float calulations: use double Internal floating point (FP) calculations should be performed using double unless there is a very good reason. This avoids headaches with conversions. Indeed, the vast majority of FP calculations were already done using double. This patch adapts most remaining calculations. Not converted where things that were based on binary representations and variables which weren't used anyway. An analysis of all instances follows: core/plannernotes.c, l.404: This was a comparison between two floats. On the left side, first an integer was cast to float then multiplied with and integer and divided by a constant double. The right hand side was an integer cast to a float. Simply divide by 1000.0 first to convert to double and continue with calculations. On the right hand side, remove the cast, because the integer will be implicitely cast to double for comparison. This conversion actually emits less instructions, because no conversion to double and back is performed. core/planner.c, l.613: Same analysis as previous case. subsurface-desktop-main.cpp, l.155: A local variable representing the version OpenGL version. Turn this into integer logic. Not only does this avoid dreaded FP rounding issues, it also works correctly for minor version > 10 (not that such a thing is to be expected anytime soon). abstractpreferenceswidget.[h/cpp]: A widget where the position is described as a float. Turn into double. desktop-widgets/divelogexportdialog.cpp, l.313: total_weight is described as float. Use double arithmetics instead. This instance fixes a truncation warning emitted by gcc.
2017-11-20 20:39:50 +00:00
return (cyl->end.mbar - prefs.reserve_gas) / 1000.0 * cyl->type.size.mliter > cyl->deco_gas_used.mliter;
else
return true;
}
/* Do a binary search for the time the ceiling is clear to ascent to target_depth.
* Minimal solution is min + 1, and the solution should be an integer multiple of stepsize.
* leap is a guess for the maximum but there is no guarantee that leap is an upper limit.
* So we always test at the upper bundary, not in the middle!
*/
static int wait_until(struct deco_state *ds, struct dive *dive, int clock, int min, int leap, int stepsize, int depth, int target_depth, int avg_depth, int bottom_time, struct gasmix gasmix, int po2, double surface_pressure, enum divemode_t divemode)
{
// When a deco stop exceeds two days, there is something wrong...
if (min >= 48 * 3600)
return 50 * 3600;
// Round min + leap up to the next multiple of stepsize
int upper = min + leap + stepsize - 1 - (min + leap - 1) % stepsize;
// Is the upper boundary too small?
if (!trial_ascent(ds, upper - clock, depth, target_depth, avg_depth, bottom_time, gasmix, po2, surface_pressure, dive, divemode))
return wait_until(ds, dive, clock, upper, leap, stepsize, depth, target_depth, avg_depth, bottom_time, gasmix, po2, surface_pressure, divemode);
if (upper - min <= stepsize)
return upper;
return wait_until(ds, dive, clock, min, leap / 2, stepsize, depth, target_depth, avg_depth, bottom_time, gasmix, po2, surface_pressure, divemode);
}
static void average_max_depth(struct diveplan *dive, int *avg_depth, int *max_depth)
{
int integral = 0;
int last_time = 0;
int last_depth = 0;
struct divedatapoint *dp = dive->dp;
*max_depth = 0;
while (dp) {
if (dp->time) {
/* Ignore gas indication samples */
integral += (dp->depth.mm + last_depth) * (dp->time - last_time) / 2;
last_time = dp->time;
last_depth = dp->depth.mm;
if (dp->depth.mm > *max_depth)
*max_depth = dp->depth.mm;
}
dp = dp->next;
}
if (last_time)
*avg_depth = integral / last_time;
else
*avg_depth = *max_depth = 0;
}
bool plan(struct deco_state *ds, struct diveplan *diveplan, struct dive *dive, int timestep, struct decostop *decostoptable, struct deco_state **cached_datap, bool is_planner, bool show_disclaimer)
First stab at simplistic dive planning This comes with absolutely no gui - so the plan literally needs to be compiled into Subsurface. Not exactly a feature, but this allowed me to focus on the planning part instead of spending time on tedious UI work. A new menu "Planner" with entry "Test Planner" calls into the hard-coded function in planner.c. There a simple dive plan can be constructed with calls to plan_add_segment(&diveplan, duration, depth at the end, fO2, pO2) Calling plan(&diveplan) does the deco calculations and creates deco stops that keep us below the ceiling (with the GFlow/high values currently configured). The stop levels used are defined at the top of planner.c in the stoplevels array - there is no need to do the traditional multiples of 3m or anything like that. The dive including the ascents and deco stops all the way to the surface is completed and then added as simulated dive to the end of the divelist (I guess we could automatically select it later) and can be viewed. This is crude but shows the direction we can go with this. Envision a nice UI that allows you to simply enter the segments and pick the desired stops. What is missing is the ability to give the algorithm additional gases that it can use during the deco phase - right now it simply keeps using the last gas used in the diveplan. All that said, there are clear bugs here - and sadly they seem to be in the deco calculations, as with the example given the ceiling that is calculated makes no sense. When displayed in smooth mode it has very strange jumps up and down that I wouldn't expect. For example with GF 35/75 (the default) the deco ceiling when looking at the simulated dive jumps from 16m back up to 13m around 14:10 into the dive. That seems very odd. Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2013-01-05 07:11:42 +00:00
{
int bottom_depth;
int bottom_gi;
int bottom_stopidx;
bool is_final_plan = true;
int bottom_time;
int previous_deco_time;
struct deco_state *bottom_cache = NULL;
First stab at simplistic dive planning This comes with absolutely no gui - so the plan literally needs to be compiled into Subsurface. Not exactly a feature, but this allowed me to focus on the planning part instead of spending time on tedious UI work. A new menu "Planner" with entry "Test Planner" calls into the hard-coded function in planner.c. There a simple dive plan can be constructed with calls to plan_add_segment(&diveplan, duration, depth at the end, fO2, pO2) Calling plan(&diveplan) does the deco calculations and creates deco stops that keep us below the ceiling (with the GFlow/high values currently configured). The stop levels used are defined at the top of planner.c in the stoplevels array - there is no need to do the traditional multiples of 3m or anything like that. The dive including the ascents and deco stops all the way to the surface is completed and then added as simulated dive to the end of the divelist (I guess we could automatically select it later) and can be viewed. This is crude but shows the direction we can go with this. Envision a nice UI that allows you to simply enter the segments and pick the desired stops. What is missing is the ability to give the algorithm additional gases that it can use during the deco phase - right now it simply keeps using the last gas used in the diveplan. All that said, there are clear bugs here - and sadly they seem to be in the deco calculations, as with the example given the ceiling that is calculated makes no sense. When displayed in smooth mode it has very strange jumps up and down that I wouldn't expect. For example with GF 35/75 (the default) the deco ceiling when looking at the simulated dive jumps from 16m back up to 13m around 14:10 into the dive. That seems very odd. Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2013-01-05 07:11:42 +00:00
struct sample *sample;
int po2;
int transitiontime, gi;
int current_cylinder, stop_cylinder;
int stopidx;
int depth;
struct gaschanges *gaschanges = NULL;
int gaschangenr;
int *decostoplevels;
int decostoplevelcount;
int *stoplevels = NULL;
Improvement for various bits of the planner Rewrite of the actual planner logic. Now ascend to the next potential stop depth. There the state is cached and we try to ascend to the next stop depth. If we hit the ceiling while doing that we go back to the cached state and wait there for a minute. Then we try again. Then loop. Converted all depth related variables from unsigned int to int. During planning, in a time step the current depth can temporarily be negative and comparisons of a negative int with an unsigned it have not the result I expected ( (int) -2 < (unsigned int) 3 turns out to be false). And we don’t really need the 32nd bit that unsigned buys us for depths. Deco stops are now shown in the same table as manually entered stops in boldface (I removed the second table to save screen estate). The gas shown in the table is still misleading as it means the gas used on the segment leading up to that event. The update of the profile only works partially upon changes in the list of available gases. Treatment of various gases is basically there but needs some more love. The ascent velocity is now provided by a function that takes the current depth as argument. Currently it always returns 10m/min but that will later be variable (and hopefully user configurable). The profile is not redrawn while deco is computed (avoiding an infinite recursion). The table got a new column for the duration of a segment while the old “duration” column was renamed “Runtime” to reflect what it actually shows. Currently, only the run time but not the duration are editable. All deco gases are used from the depth where their pO2 is 1.4bar. This should become more flexible. Calculation of the pressure drop in cylinders without configured volumes is suppressed. This solves a problem with the planner crashing when saving a dive where not all cylinders had been manually given a volume. [Short rant break: Treating 0/0 as air bites back at so many places. E.g. Cylinder data is initialized with memsetting the whole structures to 0. Then later suddenly this totally unconfigured cylinder is being treated as it would contain air. Maybe at some point this was a feature. But it lead to a naughty bug which took me over an hour to resolve. We should seriously reconsider this choice and better move to 209/0 being air if changing this everywhere is not too much trouble] Signed-off-by: Robert C. Helling <helling@atdotde.de> Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2014-04-17 08:54:55 +00:00
bool stopping = false;
bool pendinggaschange = false;
Improvement for various bits of the planner Rewrite of the actual planner logic. Now ascend to the next potential stop depth. There the state is cached and we try to ascend to the next stop depth. If we hit the ceiling while doing that we go back to the cached state and wait there for a minute. Then we try again. Then loop. Converted all depth related variables from unsigned int to int. During planning, in a time step the current depth can temporarily be negative and comparisons of a negative int with an unsigned it have not the result I expected ( (int) -2 < (unsigned int) 3 turns out to be false). And we don’t really need the 32nd bit that unsigned buys us for depths. Deco stops are now shown in the same table as manually entered stops in boldface (I removed the second table to save screen estate). The gas shown in the table is still misleading as it means the gas used on the segment leading up to that event. The update of the profile only works partially upon changes in the list of available gases. Treatment of various gases is basically there but needs some more love. The ascent velocity is now provided by a function that takes the current depth as argument. Currently it always returns 10m/min but that will later be variable (and hopefully user configurable). The profile is not redrawn while deco is computed (avoiding an infinite recursion). The table got a new column for the duration of a segment while the old “duration” column was renamed “Runtime” to reflect what it actually shows. Currently, only the run time but not the duration are editable. All deco gases are used from the depth where their pO2 is 1.4bar. This should become more flexible. Calculation of the pressure drop in cylinders without configured volumes is suppressed. This solves a problem with the planner crashing when saving a dive where not all cylinders had been manually given a volume. [Short rant break: Treating 0/0 as air bites back at so many places. E.g. Cylinder data is initialized with memsetting the whole structures to 0. Then later suddenly this totally unconfigured cylinder is being treated as it would contain air. Maybe at some point this was a feature. But it lead to a naughty bug which took me over an hour to resolve. We should seriously reconsider this choice and better move to 209/0 being air if changing this everywhere is not too much trouble] Signed-off-by: Robert C. Helling <helling@atdotde.de> Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2014-04-17 08:54:55 +00:00
int clock, previous_point_time;
int avg_depth, max_depth;
int last_ascend_rate;
int best_first_ascend_cylinder;
struct gasmix gas, bottom_gas;
bool o2break_next = false;
int break_cylinder = -1, breakfrom_cylinder = 0;
bool last_segment_min_switch = false;
int error = 0;
bool decodive = false;
int first_stop_depth = 0;
int laststoptime = timestep;
bool o2breaking = false;
int decostopcounter = 0;
enum divemode_t divemode = dive->dc.divemode;
First stab at simplistic dive planning This comes with absolutely no gui - so the plan literally needs to be compiled into Subsurface. Not exactly a feature, but this allowed me to focus on the planning part instead of spending time on tedious UI work. A new menu "Planner" with entry "Test Planner" calls into the hard-coded function in planner.c. There a simple dive plan can be constructed with calls to plan_add_segment(&diveplan, duration, depth at the end, fO2, pO2) Calling plan(&diveplan) does the deco calculations and creates deco stops that keep us below the ceiling (with the GFlow/high values currently configured). The stop levels used are defined at the top of planner.c in the stoplevels array - there is no need to do the traditional multiples of 3m or anything like that. The dive including the ascents and deco stops all the way to the surface is completed and then added as simulated dive to the end of the divelist (I guess we could automatically select it later) and can be viewed. This is crude but shows the direction we can go with this. Envision a nice UI that allows you to simply enter the segments and pick the desired stops. What is missing is the ability to give the algorithm additional gases that it can use during the deco phase - right now it simply keeps using the last gas used in the diveplan. All that said, there are clear bugs here - and sadly they seem to be in the deco calculations, as with the example given the ceiling that is calculated makes no sense. When displayed in smooth mode it has very strange jumps up and down that I wouldn't expect. For example with GF 35/75 (the default) the deco ceiling when looking at the simulated dive jumps from 16m back up to 13m around 14:10 into the dive. That seems very odd. Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2013-01-05 07:11:42 +00:00
set_gf(diveplan->gflow, diveplan->gfhigh);
set_vpmb_conservatism(diveplan->vpmb_conservatism);
First stab at simplistic dive planning This comes with absolutely no gui - so the plan literally needs to be compiled into Subsurface. Not exactly a feature, but this allowed me to focus on the planning part instead of spending time on tedious UI work. A new menu "Planner" with entry "Test Planner" calls into the hard-coded function in planner.c. There a simple dive plan can be constructed with calls to plan_add_segment(&diveplan, duration, depth at the end, fO2, pO2) Calling plan(&diveplan) does the deco calculations and creates deco stops that keep us below the ceiling (with the GFlow/high values currently configured). The stop levels used are defined at the top of planner.c in the stoplevels array - there is no need to do the traditional multiples of 3m or anything like that. The dive including the ascents and deco stops all the way to the surface is completed and then added as simulated dive to the end of the divelist (I guess we could automatically select it later) and can be viewed. This is crude but shows the direction we can go with this. Envision a nice UI that allows you to simply enter the segments and pick the desired stops. What is missing is the ability to give the algorithm additional gases that it can use during the deco phase - right now it simply keeps using the last gas used in the diveplan. All that said, there are clear bugs here - and sadly they seem to be in the deco calculations, as with the example given the ceiling that is calculated makes no sense. When displayed in smooth mode it has very strange jumps up and down that I wouldn't expect. For example with GF 35/75 (the default) the deco ceiling when looking at the simulated dive jumps from 16m back up to 13m around 14:10 into the dive. That seems very odd. Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2013-01-05 07:11:42 +00:00
if (!diveplan->surface_pressure)
diveplan->surface_pressure = SURFACE_PRESSURE;
dive->surface_pressure.mbar = diveplan->surface_pressure;
clear_deco(ds, dive->surface_pressure.mbar / 1000.0);
ds->max_bottom_ceiling_pressure.mbar = ds->first_ceiling_pressure.mbar = 0;
create_dive_from_plan(diveplan, dive, is_planner);
First stab at simplistic dive planning This comes with absolutely no gui - so the plan literally needs to be compiled into Subsurface. Not exactly a feature, but this allowed me to focus on the planning part instead of spending time on tedious UI work. A new menu "Planner" with entry "Test Planner" calls into the hard-coded function in planner.c. There a simple dive plan can be constructed with calls to plan_add_segment(&diveplan, duration, depth at the end, fO2, pO2) Calling plan(&diveplan) does the deco calculations and creates deco stops that keep us below the ceiling (with the GFlow/high values currently configured). The stop levels used are defined at the top of planner.c in the stoplevels array - there is no need to do the traditional multiples of 3m or anything like that. The dive including the ascents and deco stops all the way to the surface is completed and then added as simulated dive to the end of the divelist (I guess we could automatically select it later) and can be viewed. This is crude but shows the direction we can go with this. Envision a nice UI that allows you to simply enter the segments and pick the desired stops. What is missing is the ability to give the algorithm additional gases that it can use during the deco phase - right now it simply keeps using the last gas used in the diveplan. All that said, there are clear bugs here - and sadly they seem to be in the deco calculations, as with the example given the ceiling that is calculated makes no sense. When displayed in smooth mode it has very strange jumps up and down that I wouldn't expect. For example with GF 35/75 (the default) the deco ceiling when looking at the simulated dive jumps from 16m back up to 13m around 14:10 into the dive. That seems very odd. Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2013-01-05 07:11:42 +00:00
// Do we want deco stop array in metres or feet?
if (prefs.units.length == METERS ) {
decostoplevels = decostoplevels_metric;
decostoplevelcount = sizeof(decostoplevels_metric) / sizeof(int);
} else {
decostoplevels = decostoplevels_imperial;
decostoplevelcount = sizeof(decostoplevels_imperial) / sizeof(int);
}
/* If the user has selected last stop to be at 6m/20', we need to get rid of the 3m/10' stop.
* Otherwise reinstate the last stop 3m/10' stop.
*/
if (prefs.last_stop)
*(decostoplevels + 1) = 0;
else
*(decostoplevels + 1) = M_OR_FT(3,10);
/* Let's start at the last 'sample', i.e. the last manually entered waypoint. */
sample = &dive->dc.sample[dive->dc.samples - 1];
/* Keep time during the ascend */
bottom_time = clock = previous_point_time = dive->dc.sample[dive->dc.samples - 1].time.seconds;
current_cylinder = get_cylinderid_at_time(dive, &dive->dc, sample->time);
// Find the divemode at the end of the dive
const struct event *ev = NULL;
divemode = UNDEF_COMP_TYPE;
divemode = get_current_divemode(&dive->dc, bottom_time, &ev, &divemode);
gas = get_cylinder(dive, current_cylinder)->gasmix;
po2 = sample->setpoint.mbar;
depth = dive->dc.sample[dive->dc.samples - 1].depth.mm;
average_max_depth(diveplan, &avg_depth, &max_depth);
last_ascend_rate = ascent_velocity(depth, avg_depth, bottom_time);
/* if all we wanted was the dive just get us back to the surface */
if (!is_planner) {
/* Attn: for manually entered dives, we depend on the last segment having the
* same ascent rate as in fake_dc(). If you change it here, also change it there.
*/
transitiontime = lrint(depth / (double)prefs.ascratelast6m);
plan_add_segment(diveplan, transitiontime, 0, current_cylinder, po2, false, divemode);
create_dive_from_plan(diveplan, dive, is_planner);
return false;
}
Improvement for various bits of the planner Rewrite of the actual planner logic. Now ascend to the next potential stop depth. There the state is cached and we try to ascend to the next stop depth. If we hit the ceiling while doing that we go back to the cached state and wait there for a minute. Then we try again. Then loop. Converted all depth related variables from unsigned int to int. During planning, in a time step the current depth can temporarily be negative and comparisons of a negative int with an unsigned it have not the result I expected ( (int) -2 < (unsigned int) 3 turns out to be false). And we don’t really need the 32nd bit that unsigned buys us for depths. Deco stops are now shown in the same table as manually entered stops in boldface (I removed the second table to save screen estate). The gas shown in the table is still misleading as it means the gas used on the segment leading up to that event. The update of the profile only works partially upon changes in the list of available gases. Treatment of various gases is basically there but needs some more love. The ascent velocity is now provided by a function that takes the current depth as argument. Currently it always returns 10m/min but that will later be variable (and hopefully user configurable). The profile is not redrawn while deco is computed (avoiding an infinite recursion). The table got a new column for the duration of a segment while the old “duration” column was renamed “Runtime” to reflect what it actually shows. Currently, only the run time but not the duration are editable. All deco gases are used from the depth where their pO2 is 1.4bar. This should become more flexible. Calculation of the pressure drop in cylinders without configured volumes is suppressed. This solves a problem with the planner crashing when saving a dive where not all cylinders had been manually given a volume. [Short rant break: Treating 0/0 as air bites back at so many places. E.g. Cylinder data is initialized with memsetting the whole structures to 0. Then later suddenly this totally unconfigured cylinder is being treated as it would contain air. Maybe at some point this was a feature. But it lead to a naughty bug which took me over an hour to resolve. We should seriously reconsider this choice and better move to 209/0 being air if changing this everywhere is not too much trouble] Signed-off-by: Robert C. Helling <helling@atdotde.de> Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2014-04-17 08:54:55 +00:00
#if DEBUG_PLAN & 4
printf("gas %s\n", gasname(&gas));
printf("depth %5.2lfm \n", depth / 1000.0);
printf("current_cylinder %i\n", current_cylinder);
#endif
Improvement for various bits of the planner Rewrite of the actual planner logic. Now ascend to the next potential stop depth. There the state is cached and we try to ascend to the next stop depth. If we hit the ceiling while doing that we go back to the cached state and wait there for a minute. Then we try again. Then loop. Converted all depth related variables from unsigned int to int. During planning, in a time step the current depth can temporarily be negative and comparisons of a negative int with an unsigned it have not the result I expected ( (int) -2 < (unsigned int) 3 turns out to be false). And we don’t really need the 32nd bit that unsigned buys us for depths. Deco stops are now shown in the same table as manually entered stops in boldface (I removed the second table to save screen estate). The gas shown in the table is still misleading as it means the gas used on the segment leading up to that event. The update of the profile only works partially upon changes in the list of available gases. Treatment of various gases is basically there but needs some more love. The ascent velocity is now provided by a function that takes the current depth as argument. Currently it always returns 10m/min but that will later be variable (and hopefully user configurable). The profile is not redrawn while deco is computed (avoiding an infinite recursion). The table got a new column for the duration of a segment while the old “duration” column was renamed “Runtime” to reflect what it actually shows. Currently, only the run time but not the duration are editable. All deco gases are used from the depth where their pO2 is 1.4bar. This should become more flexible. Calculation of the pressure drop in cylinders without configured volumes is suppressed. This solves a problem with the planner crashing when saving a dive where not all cylinders had been manually given a volume. [Short rant break: Treating 0/0 as air bites back at so many places. E.g. Cylinder data is initialized with memsetting the whole structures to 0. Then later suddenly this totally unconfigured cylinder is being treated as it would contain air. Maybe at some point this was a feature. But it lead to a naughty bug which took me over an hour to resolve. We should seriously reconsider this choice and better move to 209/0 being air if changing this everywhere is not too much trouble] Signed-off-by: Robert C. Helling <helling@atdotde.de> Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2014-04-17 08:54:55 +00:00
best_first_ascend_cylinder = current_cylinder;
/* Find the gases available for deco */
if (divemode == CCR && !prefs.dobailout) { // Don't change gas in CCR mode
gaschanges = NULL;
gaschangenr = 0;
} else {
gaschanges = analyze_gaslist(diveplan, dive, &gaschangenr, depth, &best_first_ascend_cylinder);
}
/* Find the first potential decostopdepth above current depth */
for (stopidx = 0; stopidx < decostoplevelcount; stopidx++)
if (*(decostoplevels + stopidx) >= depth)
First stab at simplistic dive planning This comes with absolutely no gui - so the plan literally needs to be compiled into Subsurface. Not exactly a feature, but this allowed me to focus on the planning part instead of spending time on tedious UI work. A new menu "Planner" with entry "Test Planner" calls into the hard-coded function in planner.c. There a simple dive plan can be constructed with calls to plan_add_segment(&diveplan, duration, depth at the end, fO2, pO2) Calling plan(&diveplan) does the deco calculations and creates deco stops that keep us below the ceiling (with the GFlow/high values currently configured). The stop levels used are defined at the top of planner.c in the stoplevels array - there is no need to do the traditional multiples of 3m or anything like that. The dive including the ascents and deco stops all the way to the surface is completed and then added as simulated dive to the end of the divelist (I guess we could automatically select it later) and can be viewed. This is crude but shows the direction we can go with this. Envision a nice UI that allows you to simply enter the segments and pick the desired stops. What is missing is the ability to give the algorithm additional gases that it can use during the deco phase - right now it simply keeps using the last gas used in the diveplan. All that said, there are clear bugs here - and sadly they seem to be in the deco calculations, as with the example given the ceiling that is calculated makes no sense. When displayed in smooth mode it has very strange jumps up and down that I wouldn't expect. For example with GF 35/75 (the default) the deco ceiling when looking at the simulated dive jumps from 16m back up to 13m around 14:10 into the dive. That seems very odd. Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2013-01-05 07:11:42 +00:00
break;
if (stopidx > 0)
stopidx--;
/* Stoplevels are either depths of gas changes or potential deco stop depths. */
stoplevels = sort_stops(decostoplevels, stopidx + 1, gaschanges, gaschangenr);
Improvement for various bits of the planner Rewrite of the actual planner logic. Now ascend to the next potential stop depth. There the state is cached and we try to ascend to the next stop depth. If we hit the ceiling while doing that we go back to the cached state and wait there for a minute. Then we try again. Then loop. Converted all depth related variables from unsigned int to int. During planning, in a time step the current depth can temporarily be negative and comparisons of a negative int with an unsigned it have not the result I expected ( (int) -2 < (unsigned int) 3 turns out to be false). And we don’t really need the 32nd bit that unsigned buys us for depths. Deco stops are now shown in the same table as manually entered stops in boldface (I removed the second table to save screen estate). The gas shown in the table is still misleading as it means the gas used on the segment leading up to that event. The update of the profile only works partially upon changes in the list of available gases. Treatment of various gases is basically there but needs some more love. The ascent velocity is now provided by a function that takes the current depth as argument. Currently it always returns 10m/min but that will later be variable (and hopefully user configurable). The profile is not redrawn while deco is computed (avoiding an infinite recursion). The table got a new column for the duration of a segment while the old “duration” column was renamed “Runtime” to reflect what it actually shows. Currently, only the run time but not the duration are editable. All deco gases are used from the depth where their pO2 is 1.4bar. This should become more flexible. Calculation of the pressure drop in cylinders without configured volumes is suppressed. This solves a problem with the planner crashing when saving a dive where not all cylinders had been manually given a volume. [Short rant break: Treating 0/0 as air bites back at so many places. E.g. Cylinder data is initialized with memsetting the whole structures to 0. Then later suddenly this totally unconfigured cylinder is being treated as it would contain air. Maybe at some point this was a feature. But it lead to a naughty bug which took me over an hour to resolve. We should seriously reconsider this choice and better move to 209/0 being air if changing this everywhere is not too much trouble] Signed-off-by: Robert C. Helling <helling@atdotde.de> Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2014-04-17 08:54:55 +00:00
stopidx += gaschangenr;
gi = gaschangenr - 1;
/* Set tissue tolerance and initial vpmb gradient at start of ascent phase */
diveplan->surface_interval = tissue_at_end(ds, dive, cached_datap);
nuclear_regeneration(ds, clock);
vpmb_start_gradient(ds);
if (decoMode() == RECREATIONAL) {
bool safety_stop = prefs.safetystop && max_depth >= 10000;
track_ascent_gas(depth, dive, current_cylinder, avg_depth, bottom_time, safety_stop, divemode);
// How long can we stay at the current depth and still directly ascent to the surface?
do {
add_segment(ds, depth_to_bar(depth, dive),
get_cylinder(dive, current_cylinder)->gasmix,
timestep, po2, divemode, prefs.bottomsac);
update_cylinder_pressure(dive, depth, depth, timestep, prefs.bottomsac, get_cylinder(dive, current_cylinder), false, divemode);
clock += timestep;
} while (trial_ascent(ds, 0, depth, 0, avg_depth, bottom_time, get_cylinder(dive, current_cylinder)->gasmix,
po2, diveplan->surface_pressure / 1000.0, dive, divemode) &&
enough_gas(dive, current_cylinder) && clock < 6 * 3600);
// We did stay one DECOTIMESTEP too many.
// In the best of all worlds, we would roll back also the last add_segment in terms of caching deco state, but
// let's ignore that since for the eventual ascent in recreational mode, nobody looks at the ceiling anymore,
// so we don't really have to compute the deco state.
update_cylinder_pressure(dive, depth, depth, -timestep, prefs.bottomsac, get_cylinder(dive, current_cylinder), false, divemode);
clock -= timestep;
plan_add_segment(diveplan, clock - previous_point_time, depth, current_cylinder, po2, true, divemode);
previous_point_time = clock;
do {
/* Ascend to surface */
int deltad = ascent_velocity(depth, avg_depth, bottom_time) * TIMESTEP;
if (ascent_velocity(depth, avg_depth, bottom_time) != last_ascend_rate) {
plan_add_segment(diveplan, clock - previous_point_time, depth, current_cylinder, po2, false, divemode);
previous_point_time = clock;
last_ascend_rate = ascent_velocity(depth, avg_depth, bottom_time);
}
if (depth - deltad < 0)
deltad = depth;
clock += TIMESTEP;
depth -= deltad;
if (depth <= 5000 && depth >= (5000 - deltad) && safety_stop) {
plan_add_segment(diveplan, clock - previous_point_time, 5000, current_cylinder, po2, false, divemode);
previous_point_time = clock;
clock += 180;
plan_add_segment(diveplan, clock - previous_point_time, 5000, current_cylinder, po2, false, divemode);
previous_point_time = clock;
safety_stop = false;
}
} while (depth > 0);
plan_add_segment(diveplan, clock - previous_point_time, 0, current_cylinder, po2, false, divemode);
create_dive_from_plan(diveplan, dive, is_planner);
add_plan_to_notes(diveplan, dive, show_disclaimer, error);
fixup_dc_duration(&dive->dc);
free(stoplevels);
free(gaschanges);
return false;
}
Improvement for various bits of the planner Rewrite of the actual planner logic. Now ascend to the next potential stop depth. There the state is cached and we try to ascend to the next stop depth. If we hit the ceiling while doing that we go back to the cached state and wait there for a minute. Then we try again. Then loop. Converted all depth related variables from unsigned int to int. During planning, in a time step the current depth can temporarily be negative and comparisons of a negative int with an unsigned it have not the result I expected ( (int) -2 < (unsigned int) 3 turns out to be false). And we don’t really need the 32nd bit that unsigned buys us for depths. Deco stops are now shown in the same table as manually entered stops in boldface (I removed the second table to save screen estate). The gas shown in the table is still misleading as it means the gas used on the segment leading up to that event. The update of the profile only works partially upon changes in the list of available gases. Treatment of various gases is basically there but needs some more love. The ascent velocity is now provided by a function that takes the current depth as argument. Currently it always returns 10m/min but that will later be variable (and hopefully user configurable). The profile is not redrawn while deco is computed (avoiding an infinite recursion). The table got a new column for the duration of a segment while the old “duration” column was renamed “Runtime” to reflect what it actually shows. Currently, only the run time but not the duration are editable. All deco gases are used from the depth where their pO2 is 1.4bar. This should become more flexible. Calculation of the pressure drop in cylinders without configured volumes is suppressed. This solves a problem with the planner crashing when saving a dive where not all cylinders had been manually given a volume. [Short rant break: Treating 0/0 as air bites back at so many places. E.g. Cylinder data is initialized with memsetting the whole structures to 0. Then later suddenly this totally unconfigured cylinder is being treated as it would contain air. Maybe at some point this was a feature. But it lead to a naughty bug which took me over an hour to resolve. We should seriously reconsider this choice and better move to 209/0 being air if changing this everywhere is not too much trouble] Signed-off-by: Robert C. Helling <helling@atdotde.de> Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2014-04-17 08:54:55 +00:00
if (best_first_ascend_cylinder != current_cylinder) {
current_cylinder = best_first_ascend_cylinder;
gas = get_cylinder(dive, current_cylinder)->gasmix;
#if DEBUG_PLAN & 16
printf("switch to gas %d (%d/%d) @ %5.2lfm\n", best_first_ascend_cylinder,
(get_o2(&gas) + 5) / 10, (get_he(&gas) + 5) / 10, gaschanges[best_first_ascend_cylinder].depth / 1000.0);
#endif
}
// VPM-B or Buehlmann Deco
tissue_at_end(ds, dive, cached_datap);
if ((divemode == CCR || divemode == PSCR) && prefs.dobailout) {
divemode = OC;
po2 = 0;
add_segment(ds, depth_to_bar(depth, dive),
get_cylinder(dive, current_cylinder)->gasmix,
prefs.min_switch_duration, po2, divemode, prefs.bottomsac);
plan_add_segment(diveplan, prefs.min_switch_duration, depth, current_cylinder, po2, false, divemode);
clock += prefs.min_switch_duration;
last_segment_min_switch = true;
}
previous_deco_time = 100000000;
ds->deco_time = 10000000;
cache_deco_state(ds, &bottom_cache); // Lets us make several iterations
bottom_depth = depth;
bottom_gi = gi;
bottom_gas = gas;
bottom_stopidx = stopidx;
//CVA
do {
decostopcounter = 0;
is_final_plan = (decoMode() == BUEHLMANN) || (previous_deco_time - ds->deco_time < 10); // CVA time converges
if (ds->deco_time != 10000000)
vpmb_next_gradient(ds, ds->deco_time, diveplan->surface_pressure / 1000.0);
previous_deco_time = ds->deco_time;
restore_deco_state(bottom_cache, ds, true);
depth = bottom_depth;
gi = bottom_gi;
clock = previous_point_time = bottom_time;
gas = bottom_gas;
stopping = false;
decodive = false;
first_stop_depth = 0;
stopidx = bottom_stopidx;
ds->first_ceiling_pressure.mbar = depth_to_mbar(
deco_allowed_depth(tissue_tolerance_calc(ds, dive, depth_to_bar(depth, dive)),
diveplan->surface_pressure / 1000.0, dive, 1),
dive);
if (ds->max_bottom_ceiling_pressure.mbar > ds->first_ceiling_pressure.mbar)
ds->first_ceiling_pressure.mbar = ds->max_bottom_ceiling_pressure.mbar;
last_ascend_rate = ascent_velocity(depth, avg_depth, bottom_time);
/* Always prefer the best_first_ascend_cylinder if it has the right gasmix.
* Otherwise take first cylinder from list with rightgasmix */
if (same_gasmix(gas, get_cylinder(dive, best_first_ascend_cylinder)->gasmix))
current_cylinder = best_first_ascend_cylinder;
else
current_cylinder = get_gasidx(dive, gas);
if (current_cylinder == -1) {
report_error(translate("gettextFromC", "Can't find gas %s"), gasname(gas));
current_cylinder = 0;
}
reset_regression(ds);
while (1) {
/* We will break out when we hit the surface */
do {
/* Ascend to next stop depth */
int deltad = ascent_velocity(depth, avg_depth, bottom_time) * TIMESTEP;
if (ascent_velocity(depth, avg_depth, bottom_time) != last_ascend_rate) {
if (is_final_plan)
plan_add_segment(diveplan, clock - previous_point_time, depth, current_cylinder, po2, false, divemode);
previous_point_time = clock;
stopping = false;
last_ascend_rate = ascent_velocity(depth, avg_depth, bottom_time);
}
if (depth - deltad < stoplevels[stopidx])
deltad = depth - stoplevels[stopidx];
add_segment(ds, depth_to_bar(depth, dive),
get_cylinder(dive, current_cylinder)->gasmix,
TIMESTEP, po2, divemode, prefs.decosac);
last_segment_min_switch = false;
clock += TIMESTEP;
depth -= deltad;
/* Print VPM-Gradient as gradient factor, this has to be done from within deco.c */
if (decodive)
ds->plot_depth = depth;
} while (depth > 0 && depth > stoplevels[stopidx]);
if (depth <= 0)
break; /* We are at the surface */
if (gi >= 0 && stoplevels[stopidx] <= gaschanges[gi].depth) {
/* We have reached a gas change.
* Record this in the dive plan */
Improvement for various bits of the planner Rewrite of the actual planner logic. Now ascend to the next potential stop depth. There the state is cached and we try to ascend to the next stop depth. If we hit the ceiling while doing that we go back to the cached state and wait there for a minute. Then we try again. Then loop. Converted all depth related variables from unsigned int to int. During planning, in a time step the current depth can temporarily be negative and comparisons of a negative int with an unsigned it have not the result I expected ( (int) -2 < (unsigned int) 3 turns out to be false). And we don’t really need the 32nd bit that unsigned buys us for depths. Deco stops are now shown in the same table as manually entered stops in boldface (I removed the second table to save screen estate). The gas shown in the table is still misleading as it means the gas used on the segment leading up to that event. The update of the profile only works partially upon changes in the list of available gases. Treatment of various gases is basically there but needs some more love. The ascent velocity is now provided by a function that takes the current depth as argument. Currently it always returns 10m/min but that will later be variable (and hopefully user configurable). The profile is not redrawn while deco is computed (avoiding an infinite recursion). The table got a new column for the duration of a segment while the old “duration” column was renamed “Runtime” to reflect what it actually shows. Currently, only the run time but not the duration are editable. All deco gases are used from the depth where their pO2 is 1.4bar. This should become more flexible. Calculation of the pressure drop in cylinders without configured volumes is suppressed. This solves a problem with the planner crashing when saving a dive where not all cylinders had been manually given a volume. [Short rant break: Treating 0/0 as air bites back at so many places. E.g. Cylinder data is initialized with memsetting the whole structures to 0. Then later suddenly this totally unconfigured cylinder is being treated as it would contain air. Maybe at some point this was a feature. But it lead to a naughty bug which took me over an hour to resolve. We should seriously reconsider this choice and better move to 209/0 being air if changing this everywhere is not too much trouble] Signed-off-by: Robert C. Helling <helling@atdotde.de> Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2014-04-17 08:54:55 +00:00
/* Check we need to change cylinder.
* We might not if the cylinder was chosen by the user
* or user has selected only to switch only at required stops.
* If current gas is hypoxic, we want to switch asap */
if (current_cylinder != gaschanges[gi].gasidx) {
if (!prefs.switch_at_req_stop ||
!trial_ascent(ds, 0, depth, stoplevels[stopidx - 1], avg_depth, bottom_time,
get_cylinder(dive, current_cylinder)->gasmix, po2, diveplan->surface_pressure / 1000.0, dive, divemode) || get_o2(get_cylinder(dive, current_cylinder)->gasmix) < 160) {
if (is_final_plan)
plan_add_segment(diveplan, clock - previous_point_time, depth, current_cylinder, po2, false, divemode);
stopping = true;
previous_point_time = clock;
current_cylinder = gaschanges[gi].gasidx;
gas = get_cylinder(dive, current_cylinder)->gasmix;
#if DEBUG_PLAN & 16
printf("switch to gas %d (%d/%d) @ %5.2lfm\n", gaschanges[gi].gasidx,
(get_o2(&gas) + 5) / 10, (get_he(&gas) + 5) / 10, gaschanges[gi].depth / 1000.0);
#endif
/* Stop for the minimum duration to switch gas unless we switch to o2 */
if (!last_segment_min_switch && get_o2(get_cylinder(dive, current_cylinder)->gasmix) != 1000) {
add_segment(ds, depth_to_bar(depth, dive),
get_cylinder(dive, current_cylinder)->gasmix,
prefs.min_switch_duration, po2, divemode, prefs.decosac);
clock += prefs.min_switch_duration;
last_segment_min_switch = true;
}
} else {
/* The user has selected the option to switch gas only at required stops.
* Remember that we are waiting to switch gas
*/
pendinggaschange = true;
}
}
gi--;
}
--stopidx;
/* Save the current state and try to ascend to the next stopdepth */
while (1) {
/* Check if ascending to next stop is clear, go back and wait if we hit the ceiling on the way */
if (trial_ascent(ds, 0, depth, stoplevels[stopidx], avg_depth, bottom_time,
get_cylinder(dive, current_cylinder)->gasmix, po2, diveplan->surface_pressure / 1000.0, dive, divemode)) {
decostoptable[decostopcounter].depth = depth;
decostoptable[decostopcounter].time = 0;
decostopcounter++;
break; /* We did not hit the ceiling */
}
/* Add a minute of deco time and then try again */
if (!decodive) {
decodive = true;
first_stop_depth = depth;
}
if (!stopping) {
/* The last segment was an ascend segment.
* Add a waypoint for start of this deco stop */
if (is_final_plan)
plan_add_segment(diveplan, clock - previous_point_time, depth, current_cylinder, po2, false, divemode);
previous_point_time = clock;
stopping = true;
}
/* Are we waiting to switch gas?
* Occurs when the user has selected the option to switch only at required stops
*/
if (pendinggaschange) {
current_cylinder = gaschanges[gi + 1].gasidx;
gas = get_cylinder(dive, current_cylinder)->gasmix;
#if DEBUG_PLAN & 16
printf("switch to gas %d (%d/%d) @ %5.2lfm\n", gaschanges[gi + 1].gasidx,
(get_o2(&gas) + 5) / 10, (get_he(&gas) + 5) / 10, gaschanges[gi + 1].depth / 1000.0);
#endif
/* Stop for the minimum duration to switch gas unless we switch to o2 */
if (!last_segment_min_switch && get_o2(get_cylinder(dive, current_cylinder)->gasmix) != 1000) {
add_segment(ds, depth_to_bar(depth, dive),
get_cylinder(dive, current_cylinder)->gasmix,
prefs.min_switch_duration, po2, divemode, prefs.decosac);
clock += prefs.min_switch_duration;
last_segment_min_switch = true;
}
pendinggaschange = false;
}
int new_clock = wait_until(ds, dive, clock, clock, laststoptime * 2 + 1, timestep, depth, stoplevels[stopidx], avg_depth,
bottom_time, get_cylinder(dive, current_cylinder)->gasmix, po2, diveplan->surface_pressure / 1000.0, divemode);
laststoptime = new_clock - clock;
/* Finish infinite deco */
if (laststoptime >= 48 * 3600 && depth >= 6000) {
error = LONGDECO;
break;
}
o2breaking = false;
stop_cylinder = current_cylinder;
if (prefs.doo2breaks && prefs.last_stop) {
/* The backgas breaks option limits time on oxygen to 12 minutes, followed by 6 minutes on
* backgas. This could be customized if there were demand.
*/
if (break_cylinder == -1) {
if (get_o2(get_cylinder(dive, best_first_ascend_cylinder)->gasmix) <= 320)
break_cylinder = best_first_ascend_cylinder;
else
break_cylinder = 0;
}
if (get_o2(get_cylinder(dive, current_cylinder)->gasmix) == 1000) {
if (laststoptime >= 12 * 60) {
laststoptime = 12 * 60;
new_clock = clock + laststoptime;
o2breaking = true;
o2break_next = true;
breakfrom_cylinder = current_cylinder;
if (is_final_plan)
plan_add_segment(diveplan, laststoptime, depth, current_cylinder, po2, false, divemode);
previous_point_time = clock + laststoptime;
current_cylinder = break_cylinder;
gas = get_cylinder(dive, current_cylinder)->gasmix;
}
} else if (o2break_next) {
if (laststoptime >= 6 * 60) {
laststoptime = 6 * 60;
new_clock = clock + laststoptime;
o2breaking = true;
o2break_next = false;
if (is_final_plan)
plan_add_segment(diveplan, laststoptime, depth, current_cylinder, po2, false, divemode);
previous_point_time = clock + laststoptime;
current_cylinder = breakfrom_cylinder;
gas = get_cylinder(dive, current_cylinder)->gasmix;
}
}
}
add_segment(ds, depth_to_bar(depth, dive), get_cylinder(dive, stop_cylinder)->gasmix,
laststoptime, po2, divemode, prefs.decosac);
last_segment_min_switch = false;
decostoptable[decostopcounter].depth = depth;
decostoptable[decostopcounter].time = laststoptime;
++decostopcounter;
clock += laststoptime;
if (!o2breaking)
break;
}
if (stopping) {
/* Next we will ascend again. Add a waypoint if we have spend deco time */
if (is_final_plan)
plan_add_segment(diveplan, clock - previous_point_time, depth, current_cylinder, po2, false, divemode);
previous_point_time = clock;
stopping = false;
}
Improvement for various bits of the planner Rewrite of the actual planner logic. Now ascend to the next potential stop depth. There the state is cached and we try to ascend to the next stop depth. If we hit the ceiling while doing that we go back to the cached state and wait there for a minute. Then we try again. Then loop. Converted all depth related variables from unsigned int to int. During planning, in a time step the current depth can temporarily be negative and comparisons of a negative int with an unsigned it have not the result I expected ( (int) -2 < (unsigned int) 3 turns out to be false). And we don’t really need the 32nd bit that unsigned buys us for depths. Deco stops are now shown in the same table as manually entered stops in boldface (I removed the second table to save screen estate). The gas shown in the table is still misleading as it means the gas used on the segment leading up to that event. The update of the profile only works partially upon changes in the list of available gases. Treatment of various gases is basically there but needs some more love. The ascent velocity is now provided by a function that takes the current depth as argument. Currently it always returns 10m/min but that will later be variable (and hopefully user configurable). The profile is not redrawn while deco is computed (avoiding an infinite recursion). The table got a new column for the duration of a segment while the old “duration” column was renamed “Runtime” to reflect what it actually shows. Currently, only the run time but not the duration are editable. All deco gases are used from the depth where their pO2 is 1.4bar. This should become more flexible. Calculation of the pressure drop in cylinders without configured volumes is suppressed. This solves a problem with the planner crashing when saving a dive where not all cylinders had been manually given a volume. [Short rant break: Treating 0/0 as air bites back at so many places. E.g. Cylinder data is initialized with memsetting the whole structures to 0. Then later suddenly this totally unconfigured cylinder is being treated as it would contain air. Maybe at some point this was a feature. But it lead to a naughty bug which took me over an hour to resolve. We should seriously reconsider this choice and better move to 209/0 being air if changing this everywhere is not too much trouble] Signed-off-by: Robert C. Helling <helling@atdotde.de> Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2014-04-17 08:54:55 +00:00
}
/* When calculating deco_time, we should pretend the final ascent rate is always the same,
* otherwise odd things can happen, such as CVA causing the final ascent to start *later*
* if the ascent rate is slower, which is completely nonsensical.
* Assume final ascent takes 20s, which is the time taken to ascend at 9m/min from 3m */
ds->deco_time = clock - bottom_time - stoplevels[stopidx + 1] / last_ascend_rate + 20;
} while (!is_final_plan);
decostoptable[decostopcounter].depth = 0;
plan_add_segment(diveplan, clock - previous_point_time, 0, current_cylinder, po2, false, divemode);
if (decoMode() == VPMB) {
diveplan->eff_gfhigh = lrint(100.0 * regressionb(ds));
diveplan->eff_gflow = lrint(100.0 * (regressiona(ds) * first_stop_depth + regressionb(ds)));
}
if (prefs.surface_segment != 0) {
// Switch to an empty air cylinder for breathing air at the surface
// If no empty cylinder is found, keep using last deco gas
cylinder_t cyl = empty_cylinder;
cyl.cylinder_use = NOT_USED;
add_to_cylinder_table(&dive->cylinders, dive->cylinders.nr, cyl);
current_cylinder = dive->cylinders.nr - 1;
plan_add_segment(diveplan, prefs.surface_segment, 0, current_cylinder, 0, false, OC);
}
create_dive_from_plan(diveplan, dive, is_planner);
add_plan_to_notes(diveplan, dive, show_disclaimer, error);
fixup_dc_duration(&dive->dc);
free(stoplevels);
free(gaschanges);
free(bottom_cache);
return decodive;
First stab at simplistic dive planning This comes with absolutely no gui - so the plan literally needs to be compiled into Subsurface. Not exactly a feature, but this allowed me to focus on the planning part instead of spending time on tedious UI work. A new menu "Planner" with entry "Test Planner" calls into the hard-coded function in planner.c. There a simple dive plan can be constructed with calls to plan_add_segment(&diveplan, duration, depth at the end, fO2, pO2) Calling plan(&diveplan) does the deco calculations and creates deco stops that keep us below the ceiling (with the GFlow/high values currently configured). The stop levels used are defined at the top of planner.c in the stoplevels array - there is no need to do the traditional multiples of 3m or anything like that. The dive including the ascents and deco stops all the way to the surface is completed and then added as simulated dive to the end of the divelist (I guess we could automatically select it later) and can be viewed. This is crude but shows the direction we can go with this. Envision a nice UI that allows you to simply enter the segments and pick the desired stops. What is missing is the ability to give the algorithm additional gases that it can use during the deco phase - right now it simply keeps using the last gas used in the diveplan. All that said, there are clear bugs here - and sadly they seem to be in the deco calculations, as with the example given the ceiling that is calculated makes no sense. When displayed in smooth mode it has very strange jumps up and down that I wouldn't expect. For example with GF 35/75 (the default) the deco ceiling when looking at the simulated dive jumps from 16m back up to 13m around 14:10 into the dive. That seems very odd. Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2013-01-05 07:11:42 +00:00
}
/*
* Get a value in tenths (so "10.2" == 102, "9" = 90)
*
* Return negative for errors.
*/
static int get_tenths(const char *begin, const char **endp)
{
char *end;
int value = strtol(begin, &end, 10);
if (begin == end)
return -1;
value *= 10;
/* Fraction? We only look at the first digit */
if (*end == '.') {
end++;
if (!isdigit(*end))
return -1;
value += *end - '0';
do {
end++;
} while (isdigit(*end));
}
*endp = end;
return value;
}
static int get_permille(const char *begin, const char **end)
{
int value = get_tenths(begin, end);
if (value >= 0) {
/* Allow a percentage sign */
if (**end == '%')
++*end;
}
return value;
}
int validate_gas(const char *text, struct gasmix *gas)
{
int o2, he;
if (!text)
return 0;
while (isspace(*text))
text++;
if (!*text)
return 0;
if (!strcasecmp(text, translate("gettextFromC", "air"))) {
o2 = O2_IN_AIR;
he = 0;
text += strlen(translate("gettextFromC", "air"));
} else if (!strcasecmp(text, translate("gettextFromC", "oxygen"))) {
o2 = 1000;
he = 0;
text += strlen(translate("gettextFromC", "oxygen"));
} else if (!strncasecmp(text, translate("gettextFromC", "ean"), 3)) {
o2 = get_permille(text + 3, &text);
he = 0;
} else {
o2 = get_permille(text, &text);
he = 0;
if (*text == '/')
he = get_permille(text + 1, &text);
}
/* We don't want any extra crud */
while (isspace(*text))
text++;
if (*text)
return 0;
/* Validate the gas mix */
if (*text || o2 < 1 || o2 > 1000 || he < 0 || o2 + he > 1000)
return 0;
/* Let it rip */
gas->o2.permille = o2;
gas->he.permille = he;
return 1;
}
int validate_po2(const char *text, int *mbar_po2)
{
int po2;
if (!text)
return 0;
po2 = get_tenths(text, &text);
if (po2 < 0)
return 0;
while (isspace(*text))
text++;
while (isspace(*text))
text++;
if (*text)
return 0;
*mbar_po2 = po2 * 100;
return 1;
}