DESENSITIZED OPTIMAL CONTROL: THEORY AND APPLICATIONS
PhDAER Seminar
Monday, September 9, 2024, from 09:30 until 11:00 - Sala Consiglio DAER - Dipartimento di Scienze e Tecnologie Aerospaziali, Building B12, 2nd floor - Politecnico di Milano, Campus Bovisa, Via La Masa, 34, Milano (MI)
Desensitized Optimal Control (DOC) enables the formulation of optimal control problems that incorporate the optimal reference solution’s sensitivities to state perturbations into the optimization process.
This is achieved through the introduction of Lagrange multiplier-like quantities that act as additional state variables, capturing the desired sensitivity information.
The DOC approach yields an optimal control problem with additional states representing sensitivities and additional controls representing gains in the sensitivity dynamics.
By penalizing user-specified sensitivities, the reference solution becomes easier to control under feedback, improving nonlinear robustness.
The DOC approach can also improve robustness to parameter uncertainties by converting fixed parameters to state variables with trivial dynamics. With an appropriate cost index, the DOC approach reduces to covariance shaping.
The Intrinsic DOC (IDOC) variant forces DOC gains to match neighboring optimal control gains.
Speaker:
Dr. Hans Seywald has extensive expertise in trajectory optimization and artificial intelligence. He earned an MS in Applied Mathematics with a minor in Physics from TUM in 1984. From 1984 to 1986, he worked as a GNC engineer at DLR Oberpfaffenhofen. He completed his Ph.D. in Aerospace Engineering at Virginia Tech in 1990. Since then, he has been with Analytical Mechanics Associates Inc. (AMA), serving as a GNC engineer from 1990 to 1997, in management from 1997 to 2019, and returning to technical work in 2019.
29.08.2024