[Dev] Allow dim zero control

[jbcaillau]

allow to define problems without control, but with variables

@abavoil this is for you! will allow you to use OC.jl both with direct and indirect approaches

[jbcaillau]

@ocots do we already have flows for an ocp + u given either as u(t) or u(t, x) (feed-back)?

[ocots]

@ocots do we already have flows for an ocp + u given either as u(t) or u(t, x) (feed-back)?

For the moment, it is not possible to do so.

[jbcaillau]

@ocots do we already have flows for an ocp + u given either as u(t) or u(t, x) (feed-back)?

For the moment, it is not possible to do so.

control-toolbox/CTFlows.jl#134

[jbcaillau]

@ocots for u of dim 0, allow flow to integrate the state only (no costate):

f = Flow(o)
xf = f(t0, x0, tf, v) # v: variable

[jbcaillau]

@remydutto interested to work on that branch?

[jbcaillau]

@abavoil see also #678

[ocots]

See #722.