Level System Identification And Predictive Control Study

Predictive control based on liquid-lever system is one of the basic problems on study in the field of complex industrial process control. This paper, according to predictive control theory, discusses the theory and the application of the nonlinear system modeling and predictive control algorithm based on Multiple Model Approach.After reviewing the development of the nonlinear system identification and control, the task of this paper is given. Then the origin and development of Multiple Model are showed. And then the background,the present development, the tendency and problems of the predictive control are summarized briefly. After discussing the form, the principle, contents and the processes of System Identification of the SISO linearity discrete system,the theory of Multiple Model Approach is explained. Then algorithm of the appoach of modeling the multiple model muster of the liquid-lever system are researched. Aiming at the typical predictive control algorithm (Model Algorithmic Control, Dynamic Matrix Control, Generalized Predictive Control), the basic structure and theory of the predictive control are discussed. A detailed analysis including predictive model, methods of revising feedback and receding horizon optimization is made. The problems about the predictive control of nonlinear system are discussed. Then Multiple Model predictive control is researched including predictive model, methods of revising feedback and receding horizon optimization is made.In this paper, a multiple model of a liquid-lever system is built offline. The multiple model predictive controller based on multi-step predictive model is carried out. The simulation results of the local and the global of the system show that the predictive controller of nonlinear system based on multiple models is better than traditional PID controller. It shows that this control algorithm is an effective method for the predictive control of nonlinear system.The paper is summed up. Some problems are presented, including the combination of the intelligence control and multi-mdoel predictive control; the recognization of the model muster of choice ;the influence of parameters upon stability, robust and other capabilities, the robust of the predictive control system to modeling errors and disturbances, the choice of optimization strategy for effective and robust controller and so on.