Research Of PID Controller Based On Fuzzy Neural Network In Central Air-conditioning System

With the rapid development of our country’s economy and continuousimprovement of people’s living standard, the use of Central Air-conditioning System,which is an important part of intelligent building, becomes more and more common.At present, however, the traditional PID and fuzzy control algorithm is used in mostcontrollers of air conditioning system, which causes a insatisfactory control effect.Therefore, to find the more rational control strategy and improve the controlperformance of controller has important practical value of which can not only providea comfortable air-conditioning environment but also reduce the energy consumptionof air-conditioning effectively.As the central air-conditioning system is a complex system of nonlinear,time-varying and lagging, the conventional PID controller can’t achieve an idealcontrol effect. Neural network, which is developed on the basis of the simulation ofthe human brain way of thinking, have a ability of strong parallel processing,nonlinear approximation and self-learning; Fuzzy control, which is good at logicalreasoning and knowledge expression, can imitate the reasoning and decision-makingof human. Therefore, by the use of respective advantages of neural network and fuzzycontrol, this paper design a fuzzy neural network PID controller, which is acombination of fuzzy neural network and PID control. At the same time, for theproblem in fuzzy neural network PID controller, which are the difficult selection ofthe suitable central values and width values of membership functions, and the networkweights, second-order oscillating particle swarm optimization is used to optimizerelated parameters of fuzzy neural network PID controller. The simulation resultsshow that this fuzzy neural network PID controller can improve the stability andreliability of air-conditioning system.In order to improve the control effect of the controller in air-conditioning system,this project realizes fuzzy neural network PID controller by the way of adjusting increment of PID three parameters off-line on the basis of realizing the conventionalPID controller with FPAA chip. The experimental results show that this designedfuzzy neural network PID controller, whose structure is sample, has a good controlperformance for air conditioning system.