Load estimation in radial distribution systems using neural networks and fuzzy set techniques

Abstract

Estimates of the active and reactive load at certain load points, at regular intervals in the daily load cycle, is a fundamental information for operational planning and real-time control of distribution systems. The load estimation problem is a difficult one owing to the lack of accurate information regarding individual consumer load. Usually, only monthly kWh consumption, used for billing purposes, is available. This paper describes the conceptual basis and preliminary results of a load estimation project current under way based on actual data from a Brazilian distribution company. The project is based on the application of neural network and fuzzy set techniques to generate standard load curves for classes of consumers based on their monthly energy consumption and a large data base of load curves obtained from measurement campaigns. The load estimator uses these curves to adjust the individual load at the specified load points.