BP Neural Network Decoupling Of CFBB Main Steam Pressure And Bed Temperature And Improved Genetic Algorithms Control

As a very popular coal-fired boiler at home and abroad,circulating fluidized bed boiler(CFBB)has the advantages of environmental protection,energy saving,high combustion efficiency,wide fuel adaptability and good load regulation performance.Compared with pulverized coal fired boilers,the intermal structure of CFBB is more complex,material reaction is more intense,energy conversion and transmission process is more cumbersome.The conventional control scheme of pulverized coal fired boilers directly used in the automatic control system of combustion process of CFBB ean,t meet the control requirements of CFBB.Therefore,CFBB combustion process control system needs to be further studied.Firstly,the research background and significance of CFBB combustion process control system and the research status of decoupling technology and control technology in thermal power plant combustion system by scholars at home and abroad are analyzed,and then the boiler structure,production process and related control technology are summarized.Based on the actual operation data of 300 MW CFBB,a mathematical model of main steam pressure and bed temperature is established,and the coupling characteristics between main steam pressure and bed temperature are analyzed with the mathematical model.For the two controlled variables of main steam pressure and bed temperature in CFBB combustion system,BP neural network decoupler is used to decouple them,and then the controller based on improved genetic algorithm to optimize the PID parameters is used to control them.The essence of BP neural network decoupler is the combination of feedforward compensation decoupling principle and BP neural network,which not only possesses the dynamic decoupling characteristics of feedforward compensation decoupling algorithm,but also has the self-learming ability of BP neural network.It can effectively transform multiple input and multiple output(MIMO)system into several independent single input and single output(SISO)system.The design idea of the controller based on the improved genetic algorithm to optimize the PID parameters is to search the optimal parameters of the PID controller by genetic algorithm,so as to improve the control effect.At the same time,in order to ensure that the genetic algorithm has a good search quality and a good search schedule,the crossover operator and mutation operator of the genetic controller for the main steam pressure and bed temperature are respectively controlled by fuzzy control.For the mathematical model of main steam pressure and bed temperature of 300 MW CFBB under typical working conditions,the Simulink simulation block diagram is established by using MATLAB,and the BP neural network decoupling scheme and the improved genetic algorithm optimizing PID control method are simulated respectively.The simulation results show that the feed-forward compensation decoupling algorithm based on BP neural network can effectively decouple the main steam pressure and bed temperature,and the control scheme based on improved genetic algorithm parameters to optimize the PID parameters has faster response speed,smaller overshoot and better immunity to disturbance than the traditional PID control scheme.