A Robust Neural Network Based Short Time Electricity Price Prediction

Abstract

Price prediction and load forecasting is a difficult task for industries. Electricity price are varied according to load or demand of energy. In this article suggested a novel approach for load and price forecasting based on neural network with improved Polak-Rlbière-Polyak(PRP) learning approach. For training and testing purpose use Russian wholesale market. For the implementation and simulation of proposed approach use matrix laboratory (MATLAB) R2020a and high performance computing (HPC) lab. For the evaluation of proposed method use different result parameter mean absolute percentage error, mean square error and root mean square error. The proposed approach shows lower error rate as compare to different techniques proposed by different researchers in terms of MSE, RMSE and MAPE. For the proposed method MAPE value is 1.2069%.