Carbon Content Of Fly Ash Prediction In Coal-burned Boiler And Research Of Optimization Of Combustion

With the rapid developing of modern industry and economy,issues like energy conservation and environment protection are concerned increasingly.The coal-burning boilers used in thermal power generation not only consume large number of energy,but also cause a serious pollution problem to the environment.Therefore,to utilize the energy efficiently is an effective measure to solve those problem.The carbon content of fly ash is an important index in measuring the efficiency of boiler combustion.Currently,the measure of carbon content monitoring in most power generation plants is manual sampling and chemical analyzing in laboratory.But this mean is ineffective and costs too much manual labor.So it can’t reach the demands in practical production.In addition,there is no complete and effective method to induct the combustion of coal-burning boiler.The optimization and modification of boiler combustion are by the experience of workers.So it is hard to get the ideal running effect and leads to the waste of energy.Therefore,research on this topic has important meaning of engineering.This paper introduces the process characteristics of boiler combustion,the research state of measurement in carbon content of fly ash and optimization in boiler combustion.It also analyzes the facts that influence the carbon content of fly ash deeply.On this basis,this paper applies improved BP neural network to build model for carbon content prediction.And the wolf algorithm is applied to optimize the running conditions.At last,the simulation result verifies the validity of this method and proves that this method can induct production in reality.The main researches are as follows:1)This paper introduces the process of coal-burning boiler combustion,and analyzes how different operation conditions influence the carbon content of fly ash,this paper also introduces the means of decreasing the carbon content of fly ash in power station.2)This paper improves the error function of neural network aim to solve the problem existed in measuring carbon content in fly ash by BP neural network,and verifies the prediction effort of samples with disturbances.3)Apply the PCA to reduce the input numbers of neural network.Aim to solve the problem of too many input numbers of measuring carbon content of fly ash,this paper analyzes the contributions of input parameters to the output parameter to reduce the input parameters’ numbers.4)This paper introduces the coal-burning combustion optimization method based on the wolf algorithm.Based on the prediction of carbon content of fly ash by the improved neural network,this paper applies the wolf algorithm to optimize the operation conditions of combustion to choose the most advantage scheme for combustion.At last,this paper verifies the effort of this method by simulation.