Identification of Power Stripping Resources with Fuzzy Cluster Dynamic Approach (Case Study: West Azerbaijan Province)

Reducing electric power theft is a significant part of the potential benefits of implementing the concept of smart grid. This paper proposes a data-based approach to identify locations with unusual electricity consumption. The new distance-based method classifies the new data as violator costumers, if their distance is long to the primary consumption data. The proposed algorithm determines the number of final clusters based on the number of initial clusters and a specific index. This method is based on the dynamic fuzzy clustering algorithm. The framework removes the defect of violator subscribers by mistake. In this article are studied regional and subscriber sections of Urmia city between 2000 and 2017 years and significantly more effective than other methods of uncontrolled learning and introduced as an efficient unsupervised developed method