Study On The Control Of Mould Online Based On BP Neural Network

In order to accurately control the process of the mould,establishing the model of fuzzy control system efficiently,and developing the Baume degree measurement model that based on BP neural network.For satisfying the requirement of real-time control strategy and good performance,it is necessary to study quick learning global algorithm;through the analysis of physical and chemical reaction crystallization process,the determined factors about affecting the Baume degrees forecast can be known.The neural network model can be trained with the data about purity of KCL,temperature,current and historical data of the centrifuge.Through the analysis of the mean square deviation of the actual data and the simulated prediction with neural network,showing that the model can accurately predict and gauge the Baume degree and improving the precision of the mould finally.Firstly,this paper introduces the environment of crystallizer,the mould structure and control model,then constructs a crystallizer modeling,and introduces the application of BP neural network technology in the crystallizer control.Through the research of KCL production process,several methods of extracting KCL camallite can be introduced.This paper focus on the research and design of salt chemical industry mould level control system,building the mould level control system based on the fuzzy control strategy,accomplishing liquid level optimization of PID,achieving the corresponding system simulation research.Lastly,This dissertation is based on BP neural network control strategy,mainly introducing the neural network control strategy and algorithm,besides making the simulation of BP network prediction of Baume degree.