Research On Temperature Control Of PID Based On Fuzzy Neural Network

As one of the four great inventions,papermaking has experienced thousands of years of development.With the rapid development of modern industry and economy,as one of the national economic foundations,the paper industry is in a new stage of reform.In the context of the new era,it is required that the paper industry should not only have stable development,but also develop in a more intelligent,green,and environmentally friendly manner.This means that there is a need to continuously improve the accuracy of pulp and paper process control.However,as a typical mass transfer heat transfer process,pulp and papermaking process has many complex characteristics,such as time-varying,large lag,multiple disturbances,nonlinearity and uncertain mathematical models.So,the conventional PIDs can hardly meet control requirements.With the development of intelligent control,fuzzy control and neural network gradually recognized by people.Fuzzy control and neural network can simulate human intelligent behavior.First,fuzzy control has a strong ability of logical reasoning and knowledge expression,also,it can imitate human reasoning and decision making process.The neural network is based on the structure and characteristics of the human brain.So,neural network can simulate the human brain's way of thinking,with strong self-learning ability and parallel processing ability.Therefore,the advantages of fuzzy control and neural networks are combined to form an intelligent algorithm called fuzzy neural network.This paper applies fuzzy neural network to pulp and paper process control to improve the accuracy of conventional PID control.Three typical temperature control points are selected as the controlled objects,they are the temperature control of the replacement cooking pot in the pulping section,the temperature control of the drying section of the paper machine drying section in the papermaking section,and the causticizing temperature control in the alkali recovery section.Combination of fuzzy neural network and conventional PID to control these three typical temperature objects,and build a simulation model in Simulink.The results show that fuzzy neural network PID control has stronger anti-interference ability than the conventional PID control and Fuzzy PID,also,it has better control effect and stronger robustness.In addition,in order to verify the feasibility of fuzzy neural network PID control in practical industry,this algorithm was applied in THJSK-1 experimental platform,and implement the control algorithm on the industrial PC by the MCGS configuration software.The results show that the fuzzy neural network PID control has better dynamic performance than conventional PID control,also,it shows that this algorithm has certain feasibility in industrial practice.