Evolutionary synthesis of controller for stabilization of synthesized chaotic system oscillations

Abstract

The paper deals with a synthesis of simple feedback control law with blackbox approach for the stabilization of discrete chaotic system, both synthesized by means of analytic programming. The novelty of the approach is that a tool for symbolic regression analytic programming - is used for the purpose of stabilization of higher periodic orbits - oscillations between several values of chaotic system. The paper consists of description of analytic programming as well as a used chaotic system and detailed proposal of blackbox approach based cost function used in optimization process. Analytic programming with Self-Organizing Migrating Algorithm and Differential evolution as second algorithm for meta-evolution were used for experiments in this case.