Meteorological Parameters Influence for Medium Term Load Forecasting

The present paper reports a correction algorithm based on meteorological parameters, to apply to a medium-term load forecasting (MTLF) available system. To this aim a correlation study between various meteorological data and electric load of a Municipal Utility has been performed. In particular the analyzed meteorological data concern temperature and humidity gathered along a period often years. This correlation analysis made possible to heuristically identify the correction algorithm that is finally tested by evaluating forecasting accuracy of a MTLF system, based on a artificial neural network (ANN), using only electric time series