An electricity data cluster analysis method based on SAGA-FCM algorithm
Yang Zhou, Xiqiao Lin, Wenqian Jiang, Gang Li · 2017
The key technology to analyzing electricity data is cluster methods, of which the traditional way has already lost its agility and quality due to the increasing data volume. To this end, this paper presented an electricity data mining structure: first the higher dimensional data should be reduced to lower ones, second the reduced-dimensional results should be classified into typical usage behavior using cluster methods. Conventional cluster methods cannot deal with large-scale data sets for its slow convergence and low accuracy. This paper proposed SAGA-FCM algorithm to improve the data processing results, which is a combination of Simulated Annealing, Generic algorithm and FCM (Fuzzy C Mean) algorithm. Two examples have been made to verify the algorithm: one is to prove its availability and the other is to compare its efficiency to conventional algorithm.