Research On Equipment Group Portrait Oriented To Process Index

The application of metal aluminum is very extensive in the world today,and the demand for aluminum is also very huge.At present,the method of producing aluminum in industry is electrolytic method.By concentrating the raw materials in the electrolytic cell,electrolysis is carried out at a certain temperature.Aluminum electrolysis is a very resource-consuming production process.Improving production efficiency can save resources and protect the environment to a great extent.However,due to the complexity of the production process,the production efficiency of electrolytic aluminum still has potential for further improvement.In this paper,the electrolytic cell data is analyzed by means of portrait technology,clustering method,data mining and other methods to generate a group portrait of the electrolytic cell,depict the characteristics of the electrolytic cell,and explore the impact of different production indicators on production efficiency,with a view to providing reference for practical applications.The main research contents of this paper are as follows:Research on the clustering method of electrolytic cell equipment: the traditional single clustering method is based on all features to cluster the target,which can not fully explore the influence of certain indicators on the results.Considering the complexity of the electrolytic aluminum production environment and the mutual influence of various indicators,the clustering of the single cluster method on the electrolytic cell production data is somewhat inadequate.In this paper,the FLOC biclustering method based on row and column simultaneous clustering is adopted,and the butterfly optimization algorithm is used to optimize the initial stage of the FLOC algorithm to obtain excellent biclustering seeds;In addition,the two objective functions of the algorithm are combined so that the mean square residual error and volume of the biclusters are considered at the same time during the iteration,so that the algorithm can finally get better biclusters;At the same time,aiming at the shortcomings of butterfly algorithm that is easy to fall into local optimum and does not guarantee convergence to global optimum,a butterfly optimization algorithm integrating quantum update position mode is proposed,and other strategies such as chaotic mapping and Cauchy mutation are combined to optimize the butterfly algorithm.The optimization ability is tested on several standard test functions.The experimental results show that the improved butterfly algorithm has stronger optimization ability.The biclustering algorithm combined with the improved butterfly algorithm is also mined on the open data set.The experimental results show that the improved FLOC algorithm can find biclustering with larger volume within the acceptable threshold of mean square residual,and the clustering results are better.Research on equipment group portrait model: This paper applies the portrait technology to the aluminum electrolysis industry.Through the analysis of data,it studies the label extraction method and composition of the electrolytic cell group portrait,and combines the application of the portrait technology in other fields to study the construction method and construction process of the electrolytic cell group portrait.From data cleaning,feature analysis,tag extraction,equipment clustering,data conversion to tags,to the final portrait integration and visualization,combined with machine learning methods to mine the information in the data,the whole process is systematized and process-oriented,which can provide reference for subsequent research.By comparing and analyzing the characteristics of portraits of different groups and exploring the characteristics of electrolytic cells under different production conditions,experience can be accumulated for production.Finally,based on the actual production data,this paper uses the method proposed in the paper to generate the portrait of electrolytic cell group,and briefly analyzes the characteristics of the picture.