The Simulation Study Of GPC Algorithm Based On BSTR

Due to the difficulty of the batch reaction.It is not easy to accomplish the control of the batch reaction. General PE3 cann't accomplish high level controlling purpose. It cann't satisfy the higher and higher quality requirement of productions.With the development of computer and micro-electronics technology.Many new technologies were introduced to the controlling field.Generalized Predictive Control (GPC) is an advanced controlling method which has developped in recent years. GPC has applied in so many industres successfully.It shows nice controlling capability and robustness when the parameters and the delay-time are not certain.Furthermore,GPC developped from generalized least variance.The research is circumfused with self-adaptted control theory. It is easier than other self-adaptted control. This article combines the basic GPC algorithm and numerical value analysis and puts forward an advanced algorithm.Base on the batch stired tank reactor (BSTR) of our laboratory. Establish the model of the loop of the BSTR temperature and the flux of the cooling water. Introduce the Least Square algorithm. This article has acquired the CARIMA model of the BSTR. Furthermore, the article has acquired the simulation of GPC algorithm and improved GPC algorithm with the software called Matlab. Contrast with the PID algorithm.The GPC is better.The input of the BSTR is constrained. At last, studied the constrained generalized predictive control and get the simulation of the constrained generalized predictive control algorithm which is the theory foundation of applying in practical.