Prediction of LV substation load curves with fuzzy inference systems

This paper reports the building of a system for predicting load curves at low voltage substations supplying a set of consumers of different types (residential, industrial, etc.). The system is built by the aggregation of individual fuzzy inference systems of the Takagi-Sugeno type. The paper describes how actual measurements formed a base of raw data and how test and training sets could be built from this base. Results produced by the prediction system and their comparison with actual load curves confirm the good performance of the model