This presentation describes the recently introduced concept of Desensitized
Optimal Control (DOC), an approach that enables the design of optimal
trajectories that are insensitive to state perturbations that may be encountered
at any time along the solution. Perturbations in fixed constants can be treated
with this approach through a simple transformation that raises the status of the
relevant parameters to that of states. Thus, the DOC concept provides a method
to achieve robustness with respect to parameter uncertainties in the solution to
nonlinear optimal control problems.
As a numerical example, a Zermelo-type boat path optimization problem is treated, with uncertainties in the strength of the water current. Numerical solutions are generated with both, variational and direct optimization approaches.