Enhanced Hybrid Model for Electricity Load Forecast through Artificial Neural Network and Jaya Algorithm

This paper presents a comparative analysis of electricity load forecast for short term loads. Electricity load forecast is a method of predicting or estimating electricity load by means of historical data. It relies upon hourly day-ahead loads, given temperature forecasts, holiday information and historical loads. However, artificial neural network (ANN) based techniques are widely used with these parameters in recent years. Still, there is a scope to introduce better forecast model for enhancing the accuracy. In this paper, we have implemented a hybrid model based on ANN and Jaya optimization algorithm for finding better solutions and enhance the accuracy of the forecaster. Here, we have analyzed that, learning strength of neural network and optimizing strength of Jaya algorithm when combined generates optimal solution. To demonstrate the performance of our proposed model, we have trained our models with hourly data from the NEPOOL region (courtesy ISO New England) from 2004 to 2007 and tested on out-of-sample data from 2008 and 2009. It is observed from the experiment that hybrid model based on ANN (artificial neural network) and the Jaya optimization algorithm outperforms other comparable hybrid models in terms accuracy.