Aluminum Electrolytic General Data Recording And Analysis Of Research And Development System

Aluminum electrolysis process will produce large amounts of data, single-slot single parameter data, also have single-slot multicast data, at the same time as production operation, will not regularly add some process parameters of testing and analysis. Therefore developed aluminum production data analysis and general appear particularly important.Through the analysis of the characteristics of aluminum electrolytic data, this paper first puts forward to customize input data attributes. Then use multi-dimensional analysis technology of input data from multiple dimension analysis. Finally, through the clustering algorithm, based on the analysis of the cell in a period of similarity of the mining.The main contents and innovative as follows:1,Input data attributes. Research found that aluminum production data can be divided into a single data collection and data of two categories, and data can be in electrolytic cell can also is the points. This system was designed and the data table data attribute table, the data attributes list management data table, through the dynamic data attributes additions to achieve project process parameters, and the general ways to simulate the electrolytic profile of data entry custom most stronghold.2,For input data multi-dimensional analyzed. The design of the surrounding cell such as time, place, such units for policymakers to analyze dimensions, observation data, and with all sorts of figure and table shows the result more accurate and convenient customer master data, the relationship between them.3,Input data to clustering analyzed. Assorts Based on the analysis of the data, combined with the characteristics of production, using a K - means clustering algorithm. And as for electrolytic anode current distribution characteristics of data, using the multi-point clustering algorithm to dozens of data aggregation point for K reference points and will provide the cluster analysis, the multi-dimensional analysis of data, users in the development and evolution of the data. At the same time, will the numeric aluminum data into a nominal variable data, using the nominal variables based on clustering method analysis, show the inherent law of the production data.