Cuckoo Algorithm-BP Neural Network Based Boiler Combustion Modeling And Simulation

Data and statistics show that domestic electricity consumption is growing year by year,and from the grid structure,thermal power occupies the dominant position in the power generation industry.For thermal power plants,the high price of coal and the control of environmental emission targets have increased the burden of the thermal power industry.In order to meet the market development needs,thermal power plants have to continuously innovate and optimize their equipment to improve their competitiveness.Boilers,as one of the main equipment of thermal power plants,are particularly important for their operational reliability and operational efficiency.Traditional boiler combustion adjustment has certain limitations,in the process of combustion analysis and adjustment can only be adjusted in a single item,to analyze the impact of a single parameter change on the boiler combustion.Multiple adjustment analysis,it is difficult to accurately analyze the final change is caused by which parameter changes.Therefore,it is of great relevance and practical value to conduct research on the optimal control of boiler combustion in thermal power plants.Taking a typical coal-fired unit in China as an example,the thesis conducted a preliminary experimental study on some parameters of boiler combustion according to the actual operation of thermal power plants,and used the test data to strongly illustrate the influence of the test parameter changes on boiler combustion,and the theoretical influencing factors of boiler combustion were well verified by the test data.Based on the summary of combustion optimization adjustment,the modeling study of boiler combustion optimization was carried out using BP neural network,and the neural network model of boiler combustion was constructed.Subsequently,the cuckoo algorithm(Cuckoo Search,CS)was introduced,and the basic principles of the cuckoo algorithm and the implementation process of the algorithm were discussed.The global search capability of the Cuckoo algorithm is used to optimize the weights and thresholds of the BP neural network to improve the prediction accuracy of the neural network model of boiler combustion.At the same time,the cuckoo algorithm is used for global optimization to derive the optimal input parameters that keep the exhaust smoke temperature and fly ash carbon content within a reasonable range.The cuckoo search algorithm is used to optimize the BP neural network,which well avoids the situation that the neural network easily falls into local optimum,reduces the dependence on the initial weights and thresholds,and the cuckoo algorithm-BP neural network has higher prediction accuracy and extremely strong generalization ability.The cuckoo algorithm solves the boiler combustion optimization control parameters and obtains the optimization results of boiler combustion control,which provides a reference basis for boiler combustion adjustment.