A very short-term load forecasting in time of peak loads using interval type-2 fuzzy inference system: A case study on java bali electrical system

The process of generation, transmission and distribution of electricity to the customer must be operated properly as they are related to economic problems. One of these planning processes is a very short term forecasting. A very short term load forecasting was done one day before the day of operation that has a planning time interval every 30 minutes. Fuzzy logic is one of the methods in very short term load forecasting. This study used Interval Type-2 Fuzzy Inference System (IT-2FIS) since it has high flexibility. IT-2FIS is the development of the Footprint of Uncertainty (FOU) at IT-1FIS method that has a very flexible advantage in changing FOU, so it is supportive to form the initial processing of time series data, computation, simulation and system model validation. The implementation of IT-2 FIS was on a very short term load forecasting in peak load time. This study found an average of very short term forecasting error rate on the fourth Friday in October 2015 and 2016 is 0.71% when IT-2 FIS was implemented. This result is better than using IT-1 FIS whose average of very short term forecasting error rate is 1.11%.