UMCV Forecasting of Indian Energy Exchange Using Artificial Intelligence

Abstract

The accuracy of estimation tool is very much needed for the development of any system. The growth of electricity consumption of developing countries is more as compare to develop. Therefore, the presented work addresses a review of Indian Electricity Exchange (IEX) including efficient demand estimation technique, error evaluation for one year with all seasonal components. In this regard, three years real Unconstrained Market Clearing Volume (UMCV) data from January 2012 to December 2015 has been collected and tested for one year from January to December 2015. For the estimation, neural network (NN) model with Levenberg-Marquardt (LM) and Genetic Algorithms (GA) based NN has been trained using one year data with one moth moving window. By the experimental analysis yearly MAPE of 3.74% and 3.62% by NN and GA based NN respectively has been achieved.