Research On Configuration And Optimization Algorithm Of Calcining Process Parameters In Rotary Kiln

The development of aluminum electrolysis industry is closely related to the healthy development of China’s economy.The thesis designes to improve production efficiency,reduces waste of resources and promotes safe production through the study of the process of aluminum electrolysis production process at all stages.In the aluminum electrolysis industry,rotary kiln is a series of complex processing of petroleum coke large ring equipment.Through the research of rotary kiln process parameters and using data mining theory to analyze the influence of process parameters on the calcined coke quality parameters,it can provide aids to the scientifically understanding the process parameters of rotary kiln.The thesis studies the theoretical knowledge and related technology of aluminum electrolytic rotary kiln calcination process,it includes the calcining process of rotary kiln and the evaluation standard of rotary kiln quality parameters,learns software engineering related theory,learns data mining related knowledge,researches topics related algorithms,includes K-means clustering analysis algorithm,principal component analysis(PCA),BP neural network algorithm,particle swarm optimization algorithm and other related algorithms in the field of modern intelligence.The specific research includes:Firstly,the thesis studies the theoretical knowledge and related technology of rotary kiln calcination process to analyze the quality parameters of the rotary kiln,and selects the powder resistivity and true density of calcined coke as the object of analysis to research data mining and other related theories.The thesis uses K-means clustering analysis algorithm to analyze the rotary kiln calcination quality parameters.According to the characteristics of time correspondence in the quality parameter data and process parameter data,the classification which is analyzed by the K-means algorithm of the quality parameters corresponds to the classification of the data of process parameters,the thesis reduces the dimensions of training samples and tests samples by using Principal component analysis.Secondly,for the problem of particle swarm algorithm is easy to fall into the local extremum,the thesis optimizes particle swarm algorithm,proposes a hybrid particle swarm optimization(PSO)algorithm based on the neighboring competition.Through the fusion of BP neural network mining algorithm,DCHPSO-BP algorithm is evolved to improve the convergence efficiency and the defects of easy to sink into individual extreme value.And then it learns and trains the samples of the dimension reduction process parameters,predicts the influence of process parameters on the quality parameters of rotary kiln and establishes the prediction model for rotary kiln process parameters.Finally,using software engineering design ideas and the current popular SSH integrated architecture,the thesis builds a complete,lightweight J2EE software development quality analysis system of calcined coke in rotary kiln.The purpose of the system design is using the data mining method to predict what type of calcined coke is produced by configuring process parameters,and provides aids to the enterprise production decision.