The Improved Prediction Control Based On The Grey Dynamic Model

The Grey Prediction Control based on the grey dynamic model, is one of the effectivemethod for the complex uncertain system control. In the thesis, based on the improved greydynamic model, a novel grey prediction control which combines the fuzzy control and immunePID is put forward.The main reaearch achievements are as follows:(1) The grey prediction model often lose due effect because of the disturbance, how toexclude the role of disturbance factors is a prerequisite of using grey prediction model. Bufferoperator is one of core tools for the forecast of impact disturbance sequence.In this paper, a kindof new buffer operator with variable weight is proposed based on the principle of average tempoof time sequence and using new information. An algorithm that based on the optimal value is putforward. The method is applied in practice, the practical application shows the effectiveness ofthe proposed method.(2) In this thesis, the background value and the initial value are both improved to enhancethe precsion of GM(1,1)model, numerical example showed that the simulation precision and theprediction precision of the optimized GM(1,1)model were significantly increased.(3)A new method to improve DGM(1,1) model based on time coefficient-amended is putforward. Firstly the equal interval is amended with time coefficient and the computing method isgiven. Secondly the improved DGM(1,1) model is applied in some cases. The applicationresults show that forecast precision of the improved DGM(1,1) model has been improvedgreatly.(4)A grey predictive control algorithm with intelligent integration based on theDGM(1,1) model is put forward. The dynamic response of system is divided into several controlareas based on both current error and the change of error, and the different predictive length isgiven for different area. It synthesizes the advantages both of the fuzzy control and of the grey prediction.Simulation result shows that the algorithm features a small overshoot, quick responseand high static precision.(5) A grey predictive fuzzy immune PID conrol system based on the improvedGM(1,1)model is presented. The prediction accuracy is enhanced through the improvedbackground value and initial value.It effectively combine the grey forecast and fuzzy immunePID,the simulation results show that the proposed method has better quality and can overcomethe impact of the error and uncertainty factors.