Application of fuzzy logic to short-term planning of electric loads

The paper presents the difference between planning and forecasting in electrical power distribution networks. Fuzzy logic implementation in predicting electric loads in distribution networks for a day-long period is described. The main requirements toward forecasting of electric loads are taken into account and the following influencing factors are considered: the time (the hour t of the 24-hour load schedule, for which the forecast value is sought), the realized value of the load from the same hour of the previous day (for time t-24), and the climatic conditions (temperature). The structural scheme of the methodology of load predicting by applying fuzzy logic has been presented.The following fuzzy logic rules are applied: fuzzifying the input data and presenting them as membership functions; compiling the fuzzy rules base, making a conclusion about the estimated value of the load and applying defuzzification. The fuzzy output is converted into an exact number, corresponding to the estimated value of the load. The error in the day-ahead short-term forecasting of the electric load in a specific electrical distribution network (EDN) when applying fuzzy logic is within +2,4% to -1,7%. Conclusions are drawn from the application of fuzzy logic in forecasting the electric loads in an EDN for a day ahead.