A PID Generalzed Predictive Control Of A Kind Of Nonlinear System

Generalized predictive control algorithm has been applied to all kinds of industry control systems. Now generalized predictive control(GPC) for linearity system has been mature, but systemic and effective method for nonlinear system has not been found. The classic PID is still dominant in actual industry production process,and hinder the industry application of GPC to a certain extent.So, the paper integrates the characteristics of PID technology and GPC, presents a PID generalized predictive control algorithm for a class of nonlinear system, and improves the control quality of the system.A class of nonlinear system is replaced by a time-varying linear system after analysis, then the nonlinear system GPC that is hard to solve translates into linearity GPC that can be solved by subsistent method. Direct GPC algorithm of nonlinear system integrating online estimation algorithm of time varying parameters is studied.A PID generalized predictive control algorithm is presented by introducing PID to the traditional cost function based on nonlinear GPC theory above. A direct nonlinear PIDGPC controller is designed by orthogonal polynomials which approaches the time varying parameters of controller and RLS with forgetting factor which estimates them online.The presented algorithm is simulated by Matlab/Simulink software. The simulation results is analyzed, and the validity of the algorithm is proved.Approaching the time varying parameters by orthogonal polynomials improved the validity of parameter estimation. The direct algorithm avoids recursive solution of Diophantine equation, reduces calculation amount greatly. Introducing PID, reduces settling time, increases calculation speed, especially, when the error range is large, is more effective. The simulation results demonstrate that this method improves the efficiency and control accuracy, speeds up the tracking speed, is superior to the traditional GPC.