Short-Term Load Forecasting using Artificial Neural Network Techniques

Abstract

Artificial Neural Network (ANN) Method is applied to fore cast the short-term load for a large power system. The load has two distinct patterns: weekday and weekend-day patterns. The weekend-day pattern includes Saturdays, Sunday and Monday loads. A nonlinear load model is proposed and several structures of ANN for short term forecasting are tested. Inputs to the ANN are past loads and the output of the ANN is the load forecast for a given day. The network with one or two hidden layers is tested with various combinations of neurons, and the results are compared in terms of forecasting error. The neural network, when grouped into different load patterns, gives good load forecast.This project presents a study of short-term hourly load forecasting using Artificial Neural Networks (ANNs). To demonstrate the effectiveness of the proposed approach, publicly available data from the Australian national electricity market (NEMMCO) web site has been taken to forecast the hourly load for the Victorian power system. We predicted the hourly load demand for a full week with a high degree of accuracy. Historical load data of 2006 obtained from the NEMMCO web site was divided into several where half of them are used for training and the other half is used for testing the ANN.The inputs used were the hourly load demand for the full day (24 hours) for the state and the daily temperature, humidity and wind speeds of two major cities. The outputs obtained were the predicted hourly load demand for the next day. The neural network used has 3 layers: an input, a hidden, and an output layer. The number of inputs was 37 while the number of hidden layer neurons was varied for different performance of the network. The output layer has 24 neurons.We trained the network over 6 weeks. An absolute mean error of 2.64% was achieved when the trained network was tested on one week’s data.