Study On Soft Measurement Method Of PH Value Of Slurry In Desulfurization System

The pH value of absorber slurry is an important operating parameter of limestone-gypsum wet flue gas desulfurization system,which directly affects the economy and safety of the system.In the actual operation,the pH measuring instrument configured in the system is vulnerable to interference,often occurs blockage,failure and other phenomena,which affect the continuous and reliable operation of the system.However,these methods have disadvantages such as high cost and large time delay.Softsensing technology provides a new and effective way to measure the pH value of slurry.In this paper,the pH value of the slurry in the absorber of desulfurization system is taken as the research object,and a soft measurement model of the pH value of the slurry is established to realize the rapid and accurate measurement of the pH value of the slurry.The main research contents are as follows:First of all,the limestone-gypsum wet flue gas desulfurization system technological process and reaction mechanism were analyzed,and determine the influence factors of slurry pH value,and on the basis of the desulfurization system of field data,the related process parameters of historical data pretreatment of samples,including detecting outliers,optimize smoothing processing,normalization and sample.Secondly,the mutual information algorithm is used to screen auxiliary variables,and the maximum correlation and minimum redundancy variable selection method is used to determine the final input variable set.In the soft measurement modeling,Partial Least Squares(PLS)and Least Squares Support Vector Machine(LSSVM)were used to build the soft measurement model of grout pH value,respectively.The regularization parameters and kernel parameters of LSSVM were optimized by improved Particle Swarm Optimization(PSO).Because the static model cannot meet the requirement of prediction accuracy under the time-varying characteristics of desulfurization system,an adaptive soft sensor algorithm based on Just in Time Learning(JITL)is proposed.In the JITL framework,the LSSVM algorithm with improved PSO optimization is used as the local model,and the error correction of the model output is realized by using the deviation compensation technology to improve the stability of the model.Finally,based on the field data in the MATLAB environment for modeling simulation and experimental comparison.The results show that the prediction effect of LSSVM soft sensor model for the pH value of slurry is better than that of PLS.The improved PSO is applied to the parameter selection of LSSVM,which effectively overcomes the blindness of traditional grid search method and improves the accuracy of the model.Compared with the static model,the soft measurement model based on JITL can reflect the current state of the system,and has better prediction effect.It can effectively deal with the time-varying characteristics of the desulfurization system,and can meet the measurement requirements of pH value of the slurry on site,and provide a reference for the safety production and operation optimization of the desulfurization system.