The Application Research Of Generalized Predictive Control In The Network Control System

With the development of science and technology, the network is introduced to the control system, and then the network control system exists. However, the introduction of network in the control system changes the structure of the traditional control system and makes the analysis of the network control system become more complex. The advent of the network control system causes the great interest of the domestic and foreign scholars. So far, the study on the network control system is still a hot spot.The rapid increase of computer performance makes the application of some control algorithms based on computer become possible. Predictive control is a new type of computer control algorithm which is produced in industrial process control.This algorithm not only adopts the multistep prediction, rolling optimization and feedback emendation, but also has the advantages of good control effect, less model accuracy, strong robustness and so on. Because a lot of industrial production process has the characteristics of nonlinear uncertainties and time-varying and build a precise analytical model is very difficult, and using the classic control algorithm and modern control theory is difficult to get good control effect, the advantages of predictive control makes the method effectively use in complex industrial production process. At present, the algorithm is widely used in mechanical, metallurgy, petroleum, chemical industry and so on.In view of the peltier network control system, the artificial bee colony algorithm is used for the establishment of the system model. On this basis, use the generalized predictive control algorithm of the predictive control algorithm to design the controller.First, this paper briefly introduces the basic theory of system identification and artificial bee colony algorithm. On this basis, the artificial bee colony algorithm is used to optimize the model parameters of the network control system.Then, this paper introduces the basic principle of generalized predictive control algorithm. On this basis, design the generalized predictive controller aiming at the model I have created. When design the generalized predictive controller, find some deficiencies.At last, for the defects of the generalized predictive control algorithm, this paper respectively introduces the artificial bee colony algorithm and the stair-like control strategy to solve them.