dev-examodels

[jbcaillau]

closes #47

[github-actions]

Breakage test results
Date: 2025-05-11 23:29:58

Name	Latest	Stable
OptimalControl.jl	compat: v0.2.5	compat: v0.2.5

[jbcaillau]

@ocots two things:

• there is a version conflict for CTBase (0.15 / 0.16) detected by OptimalControl breakage: update to do?
  
• it is a bit strange to see the breakage in green in the CI while it fails... but OK
  

[ocots]

As we said before, it is green in the CI since the action passed. The result of the action is a different thing. It is not because their is a break that you don't want to merge.

[ocots]

It is normal that there is a conflict since CTParser uses CTBase 0.16 and OptimalControl 0.15.

[jbcaillau]

As we said before, it is green in the CI since the action passed. The result of the action is a different thing. It is not because their is a break that you don't want to merge.

OK

It is normal that there is a conflict since CTParser uses CTBase 0.16 and OptimalControl 0.15.

yes: is it because you wait for CTDirect to update? in this situation, breakages are useless since they will break just because of a compat issue without running any test

[jbcaillau]

@ocots please review

[ocots]

@ocots please review

I do it tonight

[ocots]

Dans l'ensemble ça me semble ok. Dur de voir les petits bugs s'il y en a. Sinon, je comprends mieux l'idée de la séparation entre ce qui est purement syntaxique du reste, lié au backend. C'est super propre.

[jbcaillau]

NB. no CUDA backend (GPU) available on github runners, removing associated test (passes on local GPU for use case no. 1)
https://github.com/control-toolbox/CTParser.jl/actions/runs/14954319794/job/42007954407?pr=51

[jbcaillau]

@amontoison

o = @def begin

            tf ∈ R, variable
            t ∈ [0, tf], time
            x = (r, v, m) ∈ R³, state
            u ∈ R, control

            x(0) == [r0, v0, m0]
            m(tf) == mf
            0.0 ≤ u(t) ≤ 1.0
            r(t) ≥ r0
            0.0 ≤ v(t) ≤ vmax

            ∂(r)(t) == v(t)
            ∂(v)(t) == -Cd * v(t)^2 * exp(-β * (r(t) - 1)) / m(t) - 1 / r(t)^2 + u(t) * Tmax / m(t)
            ∂(m)(t) == -b * Tmax * u(t)

            r(tf) → max

end

m = o(; backend = backend, grid_size = N, init = (tfs, xs, us))
s = madnlp(m)

This is MadNLP version v0.8.7, running with cuDSS v0.4.0

Number of nonzeros in constraint Jacobian............:     3804
Number of nonzeros in Lagrangian Hessian.............:    12200

Total number of variables............................:      805
                     variables with only lower bounds:      201
                variables with lower and upper bounds:      402
                     variables with only upper bounds:        0
Total number of equality constraints.................:      604
Total number of inequality constraints...............:        0
        inequality constraints with only lower bounds:        0
   inequality constraints with lower and upper bounds:        0
        inequality constraints with only upper bounds:        0

iter    objective    inf_pr   inf_du lg(mu)  ||d||  lg(rg) alpha_du alpha_pr  ls
   0 -1.2354900e+00 1.84e-01 2.00e+00  -1.0 0.00e+00    -  0.00e+00 0.00e+00   0
  ...
  21 -1.0125790e+00 6.64e-09 1.51e-08  -8.0 1.30e-02    -  1.00e+00 1.00e+00h  1

Number of Iterations....: 21

                                   (scaled)                 (unscaled)
Objective...............:  -1.0125790380561310e+00   -1.0125790380561310e+00
Dual infeasibility......:   1.5104139593169075e-08    1.5104139593169075e-08
Constraint violation....:   6.6388111075860760e-09    6.6388111075860760e-09
Complementarity.........:   9.3565895240163475e-09    9.3565895240163475e-09
Overall NLP error.......:   1.5104139593169075e-08    1.5104139593169075e-08