Title: Electricity market price forecasting using support vector machines

Authors: Yi Sun, R.C. Bansal, A.K. Bhardwaj, A.K. Srivastava

Addresses: Parasyn Controls Pty Ltd., 45 Millenium Place, Tingalpa 4173, Brisbane, Australia. ' School of Information Technology and Electrical Engineering, The University of Queensland, St. Lucia, Qld-4072, Australia. ' Department of Electrical Engineering, CET, Allahabad Agricultural Institute – Deemed University, Allahabad 211007, U.P., India. ' Department of Electrical Engineering, CET, Allahabad Agricultural Institute – Deemed University, Allahabad 211007, U.P., India

Abstract: Due to the electricity market deregulation, the techniques used for load forecasting have gradually improved over the years. Deregulation in the power system industry has caused rising requirement in planning, operating and controlling electric energy systems, which brings electricity load forecasting to a crucial level. Therefore, adequate techniques are desired for accurately predicting the load and hence assisting power companies in generating capacity scheduling, maintenance, energy planning and procurement, etc. An accurate forecast can greatly help power distribution companies to improve their electricity marketing strategies and avoid over or under unitisation of generating capacity and therefore optimises energy prices. But, to predict the load demand in real time requires a considerable amount of efforts. This paper presents a design methodology for a short term load forecasting useful for distribution companies, which are capable of interacting with users, gathering historical load data, performing a statistical analysis on the historical data and plotting graphs of the predicted load using the support vector machine (SVM). SVM is the chosen forecasting technique because many studies have concluded that SVMs produce the optimum accuracy as compared to other methods such as Naive Bayesian, but SVM has not been optimised for the domain.

Keywords: support vector machines; SVM; deregulation; short term load forecasting; STLF; electricity markets; price forecasting.

DOI: 10.1504/IJCAET.2011.037865

International Journal of Computer Aided Engineering and Technology, 2011 Vol.3 No.1, pp.1 - 18

Published online: 30 Sep 2014 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article