Demo

[jbcaillau]

@ocots To be changed into an issue once the repo exists 🤞🏾

"Journée contrôle optimal et applications" outline

FRUMAM, June 9, 2023

@jbcaillau (20')

1. rappel contexte ct: control-toolbox (bocop + hampath, interconnexion python - cf. galerie, bonus : bocop v3 / web app)
2. basic use case : fonctionnel + abstrait, direct solve)
3. goddard use case : abstrait, direct solve

@ocots (20')

1. goddard use case (suite) : indirect solve
2. CTProblems : vers un benchmark en contrôle optimal
3. architecture control-toolbox.org

[jbcaillau]

@ocots PS. I have made some minor changes in goddard, plus I suppose that we can use the F0 ⋅ g notation for the Lie operation (vector against function) in there:

g(x) = vmax - constraint(ocp, :eq2)(x, -1) # -1 for tf, not used
ub(x) = -Lie(F0, g)(x) / Lie(F1, g)(x)
μ(x, p) = H01(x, p) / Lie(F1, g)(x)

[ocots]

No it is not possible to use the dot notation since in this case we cannot tell the time and variable dependences. Another possibility would be to authorize this notation only for autonomous and fixed problems.

[jbcaillau]

WTF!?

[ocots]

Finally, I have created a repo frunam: https://github.com/control-toolbox/frunam

[jbcaillau]

👍🏽🙏🏽 just renamed it to FRUMAM 😅
https://github.com/control-toolbox/frumam

[jbcaillau]

... and everything works with julia 1.9. all set!