Clustering approach and characteristic indices for load profiles of customers using data from AMI

In this paper, the clustering approach and characteristic indices for load profiles of customers are researched. First, a method for extracting the typical electricity usage profile of an individual customer is proposed, in which the density-based DBSCAN algorithm is followed by a correction strategy aiming at eliminating the significant distortion in the extraction results caused by DBSCAN. Second, K-means clustering is performed to cluster the behavior of the customers' typical load profiles and the evaluation function for clustering effect is applied to determine the most appropriate number of clusters. Finally, the characterization indices are presented with which the profile cluster of a customer could be identified easily. The effectiveness of the proposed methods is demonstrated by the real data of power consumption of industrial and commercial customers from AMI.