Load pattern-based classification of electricity customers

Accurate knowledge of the customers' consumption patterns represents a worthwhile asset for electricity providers in competitive electricity markets. Various approaches can be used for grouping customers that exhibit similar electrical behavior into customer classes. In this paper, we focus on two approaches for customer classification-a modified follow-the-leader algorithm and the self-organizing maps. We include an overview of basic theory for these methods and discuss the performance of the customer classification on the real case of a set of customers supplied by a distribution company. We compare the results obtained from the two approaches by means of two suitably defined adequacy indicators and discuss the potential applications of the surveyed approaches.