Improved Results on Statistic Information Control With a Dynamic Neural Network Identifier

This brief proposes a novel statistic information tracking control framework for complex stochastic processes with a dynamic neural network (DNN) identifier and multiple dead zone actuators. The new driven information for the tracking problem is a series of statistic information sets (SISs) of the stochastic output signal. By using an adaptive method to adjust the weight matrices and to compensate the unknown parameters, a new control input is built with the Nussbaum gain matrix and feedback control gain. It is shown that both the identification errors of DNNs and the closed-loop SIS tracking errors converge to zero. Finally, a numerical example is included to illustrate the effectiveness of the theoretical results.