A Combined Clustering Algorithm for the Classification of Electrical Equipments Family Defects
Electrical equipments family defects are common deficiencies usually caused by some particular factors such as the material of equipment and the process of design and manufacture. In order to classify these defect data, the paper proposes a clustering algorithm combines PAM and FCM. The method uses PAM firstly to generate the cluster prototypes with the goal of lowering the initial randomness of FCM, and then it runs FCM to obtain the final clustering results. These steps are expected to ensure the accuracy of the algorithm and take less iteration. Experiments using electrical equipments family defects data-sets to test and verify the accuracy and efficiency of the algorithm are discussed. The results show that the combination methods used in this paper provide a better performance in both accuracy and run time when compared with traditional analysis approach like hierarchical clustering algorithm.
electrical equipments family defects clustering algorithm fuzzy c-means partitioning around mediod
Gang Peng Songping Tang Yun Zhang
Guangdong Power Grid Corporation Huizhou Power Supply Bureau,516000,Huizhou,Guangdong,China
国际会议
西安
英文
1-4
2016-09-01（万方平台首次上网日期，不代表论文的发表时间）