Novel statistic information control framework for non-Gaussian stochastic systems with dead-zone input

In this paper, a novel statistic information control framework is studied for non-Gaussian stochastic systems with dead-zone input. Different from both the traditional stochastic optimization objective for Gaussian systems and the PDF tracking objective for Non-Gaussian systems, the driven information and the controlled objective for control problem is the statistic information set (SIS) of the system output. Only one step neural network identification and control is considered to solved the statistic information tracking control problem. Furthermore, the stability analysis for both the identification and tracking errors is developed via the use of Lyapunov stability criterion. Computer simulations are given to demonstrate the efficiency of the proposed approach.