Research And Application Of Fuzzy Clustering Algorithm In The Analysis Of The Aluminum Electrolyzer's Status

With the development of aluminum industry's automation and information process, aluminum producers collect a lot of production data that reflects electrolyzer's operating condition, in the actual production management process, production management of electrolyzer also remain in the artificial experience, workers can not fully find and use useful information and rules that hidden in the data, resulting in a lack of scientific basis for production decisions. This article studies the fuzzy clustering algorithm thoroughly, and improves FCM algorithm, the improved FCM algorithm is applied to the aluminum electrolyzer's situation analysis, to discover information and rules that hidden in the production data,so it carries on electrolyzer's classified management.And the main content and innovation are as follows:1) Study fuzzy clustering algorithm FCM algorithm thoroughly.2) This text applies FCM clustering algorithm to the analysis of status of electrolyzer, to sort the electrolyzer, in order to get a various types of slot status.3) Combined with the character of aluminum production, we propose an improved FCM algorithm, in which we used hill-climbing algorithm to generate the initial cluster centers so that it can avoid falling into local optimal solution algorithm, and also can improve the accuracy of clustering results, and can speed up the iterative process of convergence speed.4) Using the improved FCM algorithm, we finish the design and development of the aluminum cell slot clustering analysis system,it carries on the analysis of historical slot data, and carries on contrast on FCM and improved FCM algorithm.This system achieves good results in the analysis of aluminium slot's condition, it can effectively guide the actual production, improve production efficiency,and provide the basis for the scientific management.