Kalman filter for short-term load forecasting: An hourly predictor of municipal load

The paper relies on the investigation of interconnection topic within energy management. In the competitive scenario, interconnection deals with the opportunity to achieve high quality standards in service providing. Accordingly, this work presents a short-term load forecasting (STLF) method applied to load areas of municipal or regional extension. A forecasting procedure - based on time series model using a Kalman filtering approach - has been implemented. The model provides one hour ahead forecasts of electrical load relevant to data as available for an Italian municipal utility. The results thus obtained are reported in the paper and discussed. A comparison is finally evaluated with results of a previous model, again based on Kalman filter, performing 24 hours ahead forecasts.