Optimal Control Of Boiler Combustion System Based On Fuzzy Neural Network

My country’s coal resources are very rich,accounting for about 1/3 of the world’s coal resources,which also determines that my country will be a country that uses coal as the main energy source for power generation,and thermal power still occupies half of my country’s installed power generation capacity.The boiler combustion system is one of the important systems of thermal power generation and the power source of the generator.The control level of the combustion system directly affects the safety and economy of thermal power plant production.However,the thermal combustion control system still has problems such as large inertia,nonlinearity,and strong coupling.A typical complex control system is subject to the action and interference of various control variables.When the parameters change,the entire unit changes,and at the same time brings Problems such as unstable operation and reduced thermal efficiency.Therefore,it is very important to optimize the adjustment function of the control system of the thermal power unit,improve the stability of the system,and make the control system operate with appropriate control parameters.On the basis of studying the basic working process,control task,dynamic characteristics of different subsystems and basic structure of boiler combustion system in thermal power plant,this paper takes system modeling and control system optimization as the main task to optimize the combustion control system.The main work includes the following two aspects:(1)First,excavate the historical operation data of the combustion system of a power plant,and process the historical data such as data noise reduction,filtering and initialization.Secondly,based on the controlled object of the combustion system,the control system is divided into three subsystems: the main steam pressure control system,the flue gas oxygen content control system,and the furnace pressure control system.And a mathematical model is established according to the dynamic characteristics of the controlled object.Finally,the parameters of the model transfer function are optimized by particle swarm algorithm,so as to realize the model as much as possible.Accurate and verify the accuracy of the resulting model by comparing the original data with the model output data.(2)On the basis of the above model,two control methods,PID control and conventional PID control,are used to simulate and analyze the control system respectively.Add fixed value disturbances to the original control system respectively,compare the simulation curves of the two PID control systems under disturbance,and observe the anti-disturbance capabilities of the two PID controllers;Stick test.Combustion control system is a time-varying and strongly coupled system,which makes it difficult for the control system to stabilize in the same state.Transfer the K and T parameters of the function,and simulate and analyze the system.