Font Size: a A A

Research And Development Of Anode Quality Analysis And Prediction System

Posted on:2020-04-17Degree:MasterType:Thesis
Country:ChinaCandidate:M L HaoFull Text:PDF
GTID:2381330575967965Subject:Software engineering
Abstract/Summary:PDF Full Text Request
Anode is one of the electrodes that participate in the electrochemical reaction in the aluminium electrolysis reaction.The quality of the anode also directly affects the reaction speed of the aluminium electrolysis reaction and the life of the cell.At present,the anode production mainly improves the quality of the anode from the aspects of raw material control,calcination temperature control,formulation optimization,vibration,shaping and kneading process and calcination temperature curve control.This paper establishes an algorithm model for predicting the quality of anode by studying the data of calcination process parameters and quality parameters of calcined coke,forming process parameters and the relationship between the data of calcination process parameters and the quality grade of anode,which can help technicians to adjust the production process parameters reasonably and improve the quality of anode.Firstly,the correlation between process parameters and quality parameters of different processes in anode production process is analyzed,so the characteristic parameters which have great influence on anode quality are determined,and the extraction of characteristic parameters is completed.Then,the extracted characteristic parameters are normalized.Secondly,RCNN algorithm is used to establish the prediction model of the relationship between the process parameters and quality parameters of calcination,moulding and roasting and the quality of the anode,which is used to obtain the internal relationship between the process parameters and quality parameters of calcination,moulding and roasting and the quality of the anode.Then,the time dimension is pooled by the process of anode production,and a new convolution layer and pooling layer are added to convolute the characteristic dimension of each stage to optimize the lack of missing information in RCNN model.Finally,this paper develops the anode quality analysis and prediction system by using the software development process of requirement analysis,overall design,system function design,system detailed design,coding and testing in software engineering.The system mainly realizes the analysis of the relationship between process parameters and quality parameters in different stages of production process and anode quality,and shows the predicted results of anode quality under different conditions by charts and other ways,which helps enterprises to control the configuration of main process parameters in each production process of anode production.
Keywords/Search Tags:anode quality, analysis and prediction, software engineering, recurrent neural network, convolution neural network
PDF Full Text Request
Related items