VPM-B Project - Report

Thu May 21 02:32:34 PDT 2015

About the Schreiner: yes, it is the same as Haldane on small segments.
I just want to keep it the same way as it's in the original algorithm
(also existing implementation I will test against, works on bigger
segments).

In fact most of the deco functions are declared in dive.h, when there
is something useful for both algorithms, I reuse it (mainly small
functions like plan_add_segment() etc.). But now, the main loop for
the current deco is inside planner. I think the planner should be
algorithm independent. So later, some things should be rearranged (for
better code structure).

One question: do we want to have both algorithms executed on every
diveplan or do we want to choose at some level that only VPM-B or
Buehlmann will be calculated? To be honest I've been thinking about
the second solution all the time but now it looks like the first is
better...

2015-05-21 10:49 GMT+02:00 Robert Helling <helling at atdotde.de>:
> Jan,
>
> On 21.05.2015, at 00:11, Jan Darowski <jan.darowski at gmail.com> wrote:
>
> Hi!
> Here is the current VPM-B project's state:
>
>
> Thanks a lot for your report. Here are a few comments:
>
> I've been mostly working on two things: better understanding the
> algorithm and getting through the existing subsurface planning code.
> Lately I've started coding some skeleton of the vpm-b implementation.
> My first idea was to seperate it completely from the Subsurface code.
> Robert Helling told me that there is a lot of code that can be used
> inside vpm-b and maybe I should consider just branching the existing
> code and working on it.
> The problem here is, that current implementation of the planner code
> is completely unprepared for adding new deco algorithms. I would spend
> more time on refactoring and seeking integration bugs than on writting
> vpm-b. So wanting to first focus on the algorithm, I'm creating a new
> code, which includes only few Subsurface files (dive.h, units.h).
> Doing this, I can write implementation based on the Subsurface data
> structures, using helping functions, without loosing control over the
> whole code. Of course my implementation is based on how current deco
> code is written.
>
>
> Please, please do not reinvent the wheel. There is much more that can be
> shared. All the code for computing tissue loadings (add_segment in deco.c
> and related functions) are identical in both algorithms. It is just the
> surfacing conditions (static double tissue_tolerance_calc(const struct dive
> *dive)
> ) that differ (plus an iteration in the case of VPM): For VPM, there are no
> a- and b-factors but there is a fixed allowed overpressure (that is
> determined in the “critical volume algorithm).
>
> Please reuse as much code as possible, since in the end we don’t want to
> have a separate vpm program but we want to have Buehlmann and vpm both in
> subsurface next to each other so they can be compared. Also doubling code is
> not a good idea since in later debugging this will lead to terrible problems
> when the two versions diverge.
>
>
> About the algorithm: I've been planning to prepare the first part of
> the algorithms description (what algorithm does before the deco,
> during the dive simulation) by now but it appears to be quiet trivial:
> It calculates gas pressure in the tissue compartments just like
> current deco.c implementation but using more precise Schreiner
> equation (instead of  Haldane)
>
>
> That’s what I mean. So please use the existing code (and if you find bugs
> there, correct them of course).
>
> The difference between “Haldane” and “Schreiner” are in practice
> non-exsistant: The true, underlying eqaution for both is
>
> d tissue_pressure / dt = - gamma (ambient_pressure(t) - tissue_pressure)
> (*)
>
> where for Haldane the solution with constant in time ambient pressure is
> taken while for Schreiner it can vary linearly. But what we effectively do
> is simulate the differential equation (*) since we only add segments of at
> most one minute duration for which it is pretty save to assume the ambient
> pressure to stay constant over that time. So Haldane is enough. You don’t
> gain significant precision if you take the variation of depth during that
> minute into account.
>
>
> and keeps track of the maximum crushing
> pressure (difference between gas pressure in the compartment and
> ambient pressure).
>
>
> Yes, that should be added to add_segment.
>
> And that's all!
>
> As soon as my implementation will correctly read input files and
> simulate gas pressures, I will link a repo so anyone could take a look
> at it. I hope it will be done this weekend.
>
>
> Good. Just keep in mind that in the end the input will just be for Buehlmann
> in subsurface (i.e. in terms of a struct diveplan) and not in terms of an
> input file.
>
> Best
> Robert
>
> --
> .oOo.oOo.oOo.oOo.oOo.oOo.oOo.oOo.oOo.oOo.oOo.oOo.oOo.oOo.oOo.oOo.oOo.oOo.oOo.oO
> Robert C. Helling     Elite Master Course Theoretical and Mathematical
> Physics
>                       Scientific Coordinator
>                       Ludwig Maximilians Universitaet Muenchen, Dept. Physik
>                       Phone: +49 89 2180-4523  Theresienstr. 39, rm. B339
>                       http://www.atdotde.de
>
> Enhance your privacy, use cryptography! My PGP keys have fingerprints
> A9D1 A01D 13A5 31FA 6515  BB44 0820 367C 36BC 0C1D    and
> DCED 37B6 251C 7861 270D  5613 95C7 9D32 9A8D 9B8F
>
>
>
>
>