A one-step neural network model predictive controller

MPC is a kind of computer control algorithm based on predictive model of the industry process. The classic MPCs, which are all based on linear predictive models, are unfit for the strong-nonlinearity control systems. This paper proposed a nonlinear one-step predictive model based on a two-layer BP Neural Network through consulting to the math expression of NARMAX model and constructed a one-step model predictive controller. Simulation experiment indicated that the nonlinear predictive model can excellently predict the output information of a nonlinear system, and the nonlinear model predictive controller can track several different operating points.