Sensitivity identification enhanced control strategy for nonlinear process systems

Abstract

An adaptive control strategy for nonlinear process systems is presented. The simple control law is derived based on minimizing a well-chosen performance index. The sensitivity between the controlled system input and output is necessary for implementing this strategy. The comparisons (between the adaptive controller and gain scheduling control techniques or conventional PID controller) and discussions are given to highlight their consistency, difference and the advantages of the adaptive controller. In this paper, a decomposed neural network (DNN) model is applied to the scheme. Then the stability analysis of the adaptive control system is presented based on Lyapunov theory. From the stability investigation, the stability condition for the DNN-based control system is obtained. A chemical benchmark example results show that the proposed adaptive control method can effectively control unknown nonlinear systems.