Neuro-control to energy minimization for a class of chaotic systems based on ADP algorithm

Abstract

This paper discusses the energy minimization problem of a class of chaotic systems, and constructs an optimal neuro-controller based on adaptive dynamic programming (ADP) algorithm. To learn the optimal performance index and control policy, an iterative algorithm is established. To prove the convergence of the presented iterative algorithm, theorems with rigorous and detailed proofs are given. It is proven that the iterative performance index functions are monotone decreasing and converge to the minimum energy. A simulation example is used to indicate that the presented energy minimization control method is effective. © 2013 Springer-Verlag Berlin Heidelberg.