A fuzzy expert system for peak load forecasting. Application to the Greek power system

This paper presents the development of a fuzzy expert system used to forecast the daily load curve's two minima and two maxima, for each season of the year. The inference operations of fuzzy rules are performed following the Larsen max-product implication method and the product degree of fulfillment method, while the defuzzification procedure is based on the center of area method. The proposed fuzzy expert system for peak load forecasting is tested using historical load and temperature data of the Greek interconnected power system. Test results show that the fuzzy expert system can forecast future loads with an accuracy comparable to that of neural networks, while it can also incorporate linguistic IF-THEN rules and expert's opinion.