Identifying typical load profiles using neural-fuzzy models

This paper describes a modified self-organizing algorithm, which addresses the problem of consumer classification in distribution networks according to the shape of the load profiles and the automatic extraction of the typical load profiles for each consumer category. The algorithm is a modified/weighted form of the fuzzy implementation of the Kohonen algorithm. The performances of the algorithm were studied using a set of 96 load profiles metered in the distribution network of a public utility in Romania. The algorithm produced 9 typical load profiles. The proposed approach was able to capture the quantitative and/or qualitative differences between load profiles of different consumers with same activities