Note: This content is accessible to all versions of every browser. However, this browser does not seem to support current Web standards, preventing the display of our site's design details.


Model-Free Algorithms for Stochastic Optimisation, applied to Optimal Control Problems


O. Karaca

Semester Thesis, FS15 (10433)

This semester project concentrates on three model-free control algorithms which are simultaneous perturbation stochastic approximation, robust adaptive dynamic programming for linear and nonlinear systems. The essential feature of SPSA is the requirement of only two measurements of the loss function to approximate gradient for iteratively seeking the optimal control policy. In contrast, robust adaptive dynamic programming is a novel policy iteration approach to solve algebraic Riccati equation or the HJB equation using online information of state and input. RADP is derived from different tools of modern control for both linear and nonlinear systems without requiring the apriori knowledge of the system. In this project, practical online and offline algorithms are developed and applied to several simulation schemes and real-time implementation of an inverted and self-erecting pendulum.

Supervisors: Angelos Georghiou, Paul Beuchat, John Lygeros


Type of Publication:

(13)Semester/Bachelor Thesis

J. Lygeros

File Download:

Request a copy of this publication.
(Uses JavaScript)
% Autogenerated BibTeX entry
@PhdThesis { Xxx:2015:IFA_5190
Permanent link