Static load model adjustment using fuzzy logic and differential evolution

This paper discusses the use of a hybrid system in the field of Computational Intelligence in order to refine the measurement dataset used for determination of the parameters for static load models in Electric Power Systems (EPS). The objective is reducing the effects of natural fluctuation observed in measurement dataset resulting from random loading aggregation and disaggregation on the system under study. Specifically, it is used a fuzzy logic system for processing the filtering of measurement data obtained in the field. It is also important to note that the used approach makes the adjustment of the membership functions of a fuzzy linguistic variables using a meta heuristic called differential evolution. As a result, it is possible to get parameters both for ZIP and Exponential models with mean errors lower than those obtained with raw measurement dataset. The validation of this proposal uses measures that have been made at a CEMIG utility substation in Minas Gerais state, Brazil.