Prediction and efficiency evaluation of solar energy resources by using mixed ANN and DEA approaches

Abstract

The main objective of this paper is to estimate the monthly average global solar radiation and evaluate the energy efficiency for the cities Jamshedpur, Ranchi, Dhanbad, Hazaribagh and Bokaro in the state Jharkhand using Artificial Neural Network (ANN) and Data Envelopment Analysis (DEA). ANNs are being popular in prediction of Global Solar Radiation (GSR) and DEA is popular in efficiency assessment, so this paper uses both the methods in reaching the goal. Firstly, the ANNs are used to estimate the global solar radiation and secondly, the DEA evaluates the efficiency. Using the data base of monthly average extra terrestrial solar radiation of the cities Jamshedpur, Ranchi, Dhanbad, Hazaribagh and Bokaro as input, ANN estimated the output average monthly Global Solar Radiation (GSR). After this the target for the ANN and output generated by ANN are used for the further efficiency evaluation through DEA. This paper investigates that the method considered is suitable for prediction and efficiency evaluation.