Improved short-term electricity load forecasting using extreme learning machines

Short term forecasting is an essential tool in energy companies to take the decisions about power generation, transmission and day-to-day utility operations. A number of techniques are used in the field of Short Term Load Forecasting (STLF), like statistical and artificial neural network technique. This paper proposes an extreme learning machine based STLF technique that considers relative difference in percentage of load(RDL) at different intervals as one of the main characteristics of the system load. Here for analysis, historical data are taken from Independent Electricity System Operator(IESO) for Ontario province. Forecasting results obtained by this new approach have been presented and compared with the benchmark neural network based model like NN-GA, NN-PSO and NN-ABC, which confirms its applicability in forecasting domain.