Multiple data-based ADP structures to solve the infinite horizon optimal control problem

This paper proposes the multiple actor-critic structures to obtain the optimal control via input-output data. The shunting inhibitory artificial neural network (SIANN) is used to classify the input-output data into one of several categories. Different performance index functions may be define for disparate categories. Neural networks are used to approximate the critic network and action network, respectively. It is proven that the model error asymptotically converges to zero and the closed unknown system is uniformly ultimately bounded. Simulation results demonstrate the performance of the proposed optimal control scheme for the unknown nonlinear system.