Short term load forecasting in Mauritius using Neural Network

Short term load forecast is an important requirement in power system planning and secure operation. The aim of this paper is to implement an Artificial Neural Network (ANN) method for short-term load forecasting. A first model is applied to predict the peak load of each day over a week, a second model to predict the peak load of the next day, and a third one to predict the load of the next hour. Three-layered feed-forward neural network architecture with back propagation algorithm has been proposed in each case. Power consumption data is collected from the Central Electricity Board of Mauritius for training and testing of the artificial neural network. Based on results obtained, the ANN method is a promising technique for short term load forecasting.