One Hour Ahead Load forecasting Using Support Vector Machines

Abstract

In this paper a Novel Neural Network Technique is presented using Support Vector Machine to develop a Short Term Load Forecasting (STLF) model for coming 24 hour load prediction considering various weather factors (Temperature & Humidity). In the STLF, future Load is planned up to weak ahead. Many approaches are used to develop a short term load forecasting model consisting of traditional, Artificial Intelligence (AI) and hybrid model. However there is intense need to develop a more accurate and efficient model for STLF. Detailed experiments were performed on AEMO and California Electricity market data. The results of the proposed technique have been compared with ANN models. Experimental results show that proposed technique is better in terms of accuracy, prediction & training time.