Fuzzy Based Short Term Load Forecasting and its Validation

Abstract

A fuzzy based Short-term load forecasting (STLF) in strength gadget operations is the theme of this Paper. Precise load forecasting is necessary for lowering technology fee and base load potential as it presents load forecast generation scheduling and unit dedication decisions. In this particular piece of analysis, the inputs that are utilized by the fuzzy interfacing system are the month - to - month averages in both ‘Time’ and ‘Temperature’ & monthly ‘Historical Load data’ and the output is referred to as the 'forecasted load’. A set of six triangle membership functions have been created from the input variable ‘Time’. The membership functions are Mid night, Morning,For-Noon,Afternoon,Evening,Fornight.Another input variable ‘Temperature’ has been divided into three triangular membership function. They are Low, Medium & High also we have another input is ‘Historical load data’ which is divided into three membership function & these membership function are Low Load, High Load, Very High Load. Real-world implementation studies of forecast load distribution in a one-story building, Kaliganj, Jehnaidah. Real consumption of April, October and December (month) load is compared to the fuzzy predicted load values. The expected and actual loads are quite similar & average MAPE of three months is 15.87%. Increasing the membership functions' number as well as taking the specific location temperatures all these contribute to reducing the percentage of error. Keywords: Fuzzy interfacing system, Fuzzy logic, Membership function, Triangle membership function, Defuzzification. MAPE