Evaluation of effectiveness of ANN for feature selection based electricity price forecasting

Abstract

Electricity price is one of the most concurrent aspects of power system planning. An accurate method of forecasting is required for several economic and operational advantages. In this paper effectiveness of artificial neural network (ANN) is evaluated by selecting the most influential input variable based forecasting of price in Australian electricity market using the price and total demand data of Queensland (QLD) region from January 2016 to June 2017. Using these data, monthly and weekly forecasting of electricity price is carried out. The mean absolute percentage error (MAPE) and root mean square error (RMSE) is determined by evaluating the effectiveness of proposed method. The best result shows the monthly MAPE as 1.94% and weekly MAPE as 1.06% considering only total demand as input to the ANN.