Ensemble Kalman Filtering for Inverse Optimal Control

Relatore
Hien Tran - North Carolina State University
Data e ora
mercoledì 2 maggio 2018 alle ore 14.30 - Aula M
Referente
Referente esterno
Data pubblicazione
10 aprile 2018
Dipartimento
Informatica  

Riassunto

The solution to a nonlinear optimal control problem is defined in terms of the solution to the Hamilton-Jacobi-Bellman equation, which has been solved for linear systems but is very difficult to solve for general nonlinear systems. An alternative approach is find a tabilizing feedback control first, then establish that it optimizes a specified cost functional. This is known as the inverse optimal control problem. Solving the inverse optimal control problem for discrete-time nonlinear systems requires the construction of a stabilizing feedback control law based on a control Lyapunov function (CLF). However, there are few systematic approaches available for defining appropriate CLFs.  We propose an approach that employs Bayesian filtering methodology to parameterize a quadratic CLF. In particular, we use the ensemble Kalman filter (EnKF) to estimate parameters used in defining the CLF within the control loop of the inverse optimal control problem formulation.  Using the EnKF in this setting provides a natural link between uncertainty quantification and optimal design and control, as well as a novel and intuitive way to find the one control out of an ensemble that stabilizes the system the fastest.  Results are demonstrated on both a linear and nonlinear test problem.

Contact persons: Antonio Marigonda e Marco Caliari
 





© 2002 - 2021  Universit√† degli studi di Verona
Via dell'Artigliere 8, 37129 Verona  |  P. I.V.A. 01541040232  |  C. FISCALE 93009870234