| Gray correlation clustering method is a use of the concept of grey correlationdegree to determine the degree of similarity between the various subsystems method.Fuzzy clustering method is based on fuzzy equivalence relations and fuzzy similarityrelation, cut different set for fuzzy equivalent matrix to achieve a dynamic clusteringmethod. Put fused together gray correlation clustering and fuzzy clustering method togenerate a new clustering method which is gray correlation fuzzy clustering method.This paper studies gray correlation fuzzy clustering method with fuzzy math andgray system as basis, gray correlation clustering and fuzzy clustering method as thebased approach. Firstly, according to the characteristics of gray correlation clusteringmethod and fuzzy clustering method give the gray correlation fuzzy clustering methodof implementation steps. Second, starting from the old gray correlation fuzzy clusteringmethod, put the method of entropy weights into it, weighting to the date makes theresults of sorting and clustering more scientificity and accuracy, at the same time,compare and apply it in practical problems. Finally, seven kinds of commonly usedalgorithms for structural similarity matrix has been compared and studied. When thesystem data in the same orders of magnitude, the max-min method, the minimum ofarithmetic average method, the geometric mean minimum method, the angle cosinemethod and the similarity coefficient method are better methods in constructingsimilarity coefficient, When the system data in different orders of magnitude, theabsolute value of the subtraction method and the absolute value of the reciprocalmethod are better methods in constructing similarity coefficient. On this basis, Put fusedtogether the absolute value of the reciprocal method and new gray correlation fuzzyclustering method to analysis the sorting and clustering of mainly energy production in some provinces nationwide. at the same time, this better results can use for energydepartments. |
PDF Full Text Request