抄録
In this paper, an efficient robust control method for chaotic system introducing the concept, the edge of chaos (:boundary status between chaos and non-chaos), is proposed. To realize this concept, we introduce an extended performance index which consists of two parts. One is for achievement of the system's objects, another is for keeping the system edge of chaos. Parameters of the neural network controller are adjusted to minimize the value of the extended performance index and achieve the above two objects using Random Search with Intensification and Diversification (RAS-ID). Through Lorenz chaotic system control simulation, it is clarified that our proposed method is superior to the conventional neuro control method in the view point of the generalization ability.
