Research On GEP Algorithm In Heavy Metals Speciation Prediction Modelling

Heavy metals in municipal solid waste incineration bottom ash experiences long-term complex geochemical reactions in landfills,the occurrence of morphological changes and migration is a continuing potential risk to surrounding environment.The form of heavy metals determines the migration and transformation of heavy metals in the environment,so some researchers use chemical extraction,model calculation and instrumentation to analyze the chemical form of heavy metals in the environment,and to interpret the regularities of migration and transformation of heavy metals in the environment,thereby reducing the secondary pollution hazards brought by the municipal solid waste incineration technology.However,due to the different methods used in the above research methods and different data analysis methods,resulting in no unique,incomplete and other defects in the results of the study,coupled with limited field research data and high cost of laboratory simulation test,so studying on the chemical speciation of heavy metals in the environment is extremely difficult.Therefore,the purpose of this paper is to establish a model for heavy metal speciation predicting,obtain the data of long-term migration and transformation of heavy metals in landfill,and to provide a basis for studying the long-term migration and transformation mechanism of heavy metals.With the development of computer science and technology,more and more researchers begin to apply the artificial intelligence algorithm to the establishment of prediction model.Among them,Gene Expression Programming(GEP)is a new adaptive evolutionary algorithm,which is also an artificial intelligence method for mathematical modeling.In solving function discovery,classification analysis and time series analysis,GEP shows excellent performance,it is very suitable to solve the problem of time series data prediction.Therefore,this paper mainly studies the application of GEP algorithm in the establishment of heavy metal form prediction model.This paper first highlights the main advantages and characteristics of GEP algorithm in solving complex problems by comparing with other evolutionary algorithms,and then introduces the basic components and evolution process of GEP algorithm in detail.Although the GEP algorithm has high computational efficiencyand precision compared with other evolutionary algorithms,it still has the possibility of falling into local convergence and the shortcomings of iterative computation,which affects the effect of solving practical problems.Therefore,in order to improve these two cases,the GEP algorithm is improved from the aspects of fitness function and genetic operation,and the improved GEP algorithm is applied to the prediction model of heavy metal form.It is used to predict the change regularities of heavy metal speciation in the future time points.Because the experimental data has been a simple time series,it is difficult to accurately predict the data of the heavy metal form in the landfill with the change of the future time.In order to further improve the prediction accuracy,this paper introduces the sliding window prediction method(SWPM)into the establishment of the forecasting model,and combines with the improved GEP algorithm to form the GEP sliding window prediction method(GEP-SWPM).The data are reconstructed by phase space,so as to establish a multidimensional model of heavy metal form prediction.Finally,this paper uses the standard GEP algorithm,the improved GEP algorithm and the GEP-SWPM algorithm to predict the change of heavy metal morphology.The experimental results show that the prediction accuracy of the GEP-SWPM algorithm is greatly improved compared with the former two,and it is more suitable for the prediction of the change of the form of the heavy metal.It is helpful to find the regularity of the change of the morphology of the heavy metal.Which provides a good data base for the study of the long-term migration and transformation mechanism of heavy metals.