Short term load forecasting of Eid Al Fitr holiday by using interval Type-2 Fuzzy Inference System (Case study: Electrical system of Java Bali in Indonesia)

摘要

Holiday of Eid Al Fitr causing decreased demand of electrical energy until 28.66% in case of Java Bali electricity system in Indonesia. The reduction of electrical demand needs to be known because it related to the efficiency of generation power system. Forecasting methods used in this research is IT-2 FIS. By analyzing the peak load on the day and 4 days before Eid Al Fitr in the previous year continued analysis by using IT-2 FIS will be obtained at the peak load forecasting Eid Al-Fitr in the coming year. The accuracy of this method is shown with the average of error value in 2012, 2013 and 2014 amounting to 0,2431%. This value is better than using the IT-1 FIS which has a value of error amounting to 0.3026%. The research concluded that IT-2 FIS has a value of error smaller than IT-1 FIS.