ARTIFICIAL NEURAL-NETWORK-BASED ELECTRICAL LOAD PREDICTION FOR FOOD RETAIL STORES

It has been shown by a number of investigators that artificial neural networks (ANNs) can be more reliable and effective building energy pre dictors than traditional simulation models. This paper presents the re sults from comparisons of the predictive accuracy of two commonly used neural networks employed for the prediction of the electrical load of a retail food store. The networks used were the multi-layered percept ron (MLP) and radial basis function (RBF). The MLP network was found t o perform better than the RBF network particularly in the prediction o f fluctuations of the electrical energy around the base and maximum lo ads. Further work will be carried out to optimise the structure and pr ediction accuracy of the two networks. (C) 1998 Elsevier Science Ltd. All rights reserved.