Research On Online And Offline Collaborative Operation Optimization Of Wet Desulfurization System In Coal-Fired Power Plants

With the country’s increasingly stringent requirements for environmental governance,the coal power industry has greatly reduced pollutant emissions by implementing ultra-low emissions and energy-saving technological transformation.The limestone-gypsum wet desulfurization technology is the most widely used for flue gas desulfurization in my country.At present,the operator of the desulfurization system adjusts the pump combination and the p H value of the slurry according to the outlet concentration.The operation mode is relatively extensive.While reducing SO2 emissions,the power plant’s material consumption Energy consumption is bound to increase.Under the "contradiction" between emission reduction and consumption reduction,it is of great significance to study the operation optimization of desulfurization systems in coal-fired power plants.This paper takes the desulfurization system of a 660 MW unit as the research object,formulates its operation optimization strategy based on a large amount of historical data,and adopts the method of online and offline collaborative optimization to further reduce consumption on the basis of emission reduction.First,the technological process and chemical reaction mechanism of the wet desulfurization system are briefly introduced,and the influence of various factors on the desulfurization efficiency is analyzed to provide theoretical basis for offline data mining and online intelligent optimization analysis.Secondly,based on wavelet transform,the original data is denoised and stabilized,and an incremental data mining algorithm based on evolutionary theory is proposed to perform offline calculation,retain all valid information,and divide the work of each fixed-speed pump combination according to mechanism and mathematical statistics.Regional scope,forming a desulfurization system operating condition library,from which the optimal operation strategy can be extracted according to the current operating conditions.Then,in view of the problem that offline calculation cannot cover all working conditions,an online modeling calculation method is proposed.Based on the BP neural network algorithm,the condition correction is introduced to make the trained model fit the mechanism.Based on this model,the liquid-gas ratio is predicted and converted.Circulating the slurry volume,establishing an objective function for the energy consumption of the slurry circulating pump,using the moth to fire algorithm to calculate the combination of the fixed speed pump and the frequency of the variable frequency pump with the lowest energy consumption,and traversing the calculation according to the p H value range of the slurry to obtain the slurry p H with the lowest comprehensive cost.value,fixed speed pump combination and variable frequency pump frequency.Finally,the long-short-term memory neural network is used to predict the boundary conditions of the desulfurization system,and the operating condition database is self-filled according to the desulfurization mechanism and dimensionality reduction mechanism.In order to search the optimal strategy to meet the current working conditions in the working condition library,it can meet the operability requirements of the desulfurization system and reduce the energy consumption.Research and field use show that this method can effectively reduce energy consumption and improve the economic benefits of coal-fired power plants,and has certain practicability.