Optimal Chaos Control through reinforcement learning

Gadaleta, Sabino , Gerhard Dangelmayr
Optimal Chaos Control through reinforcement learning
Chaos, 9, 775, 1999

Abstract: A general purpose chaos control algorithm based on reinforcement learning is introduced and applied to the control of chaotic systems and to the targeting problem. The algorithm does not require any information about the dynamical system or the location of periodic orbits. The computationla complexity of the problem is reduced through approximation of the phase space by means of the Neural-Gas vector quantization technique.