A nonlinear model predictive control based on pseudolinear neural networks

Abstract

A nonlinear model predictive control based on pseudolinear neural network (PNN) is proposed, in which the second order based optimization is adopted. The recursive computation of Jacobian matrix is also proposed. The stability analysis of the closed loop model predictive control system is presented based on Lyapunov theory. From the stability investigation, the sufficient condition for the asymptotic stability of the neural predictive control system is obtained. The simulated example of the continuous stirred tank reactor (CSTR) illustrated the satisfactory result based on the proposed control strategy in this paper.