1 Short Term Load Forecasting Using Predictive Modular Neural Networks V. Petridis, A. Kehagias, A. Bakirtzis and S. Kiartzis Aristotle University of Thessaloniki, Greece email: petridis@vergina.eng.auth.gr 1 Abstract In this paper we present an application of predictive modular neural networks (PREMONN) to short term load forecasting. PREMONNs are a family of probabilistically motivated algorithms which can be used for time series prediction, classification and identification. PREMONNs utilize local predictors of several types (e.g. linear predictors or artificial neural networks) and produce a final prediction which is a weighted combination of the local predictions; the weights can be interpreted as Bayesian posterior probabilities and are computed online. The method is applied to short term load forecasting for the Greek Public Power Corporation dispatching center of Crete, where PREMONN outperforms conventional prediction techniques. 2 Problem Formulation We are given a sequence y t , t=1,2, ... , where (for each t) y t has dimensions 24× 1; each of the y t components corresponds to the load of a particular hour of the day on day no. t. The predictors have the general form y t =f(y t-1 , y t-2 ,..., y t-N ), in other words one may use data from N days from the past load history. At midnight of day no. t -1 it is required to provide a prediction for the 24 hours of day t. This prediction will be used for scheduling the power generators to be activated in the following working day. Typical load for a winter and a summer day are presented in Picture 1. 0 100 200 1 6 11 16 21 Hour J a n . 1 s t, 1 9 9 3 July 1st, 1993 Picture 1: Two representative daily loads.