Adaptive Short-Term Forecasting of Hourly Loads Using Weather Information

A comprehensive computer-oriented procedure for Probablistic forecasting of hourly power-system loads with lead times of 1 to 24 hours is described. The forecasts produced are based on both historical load data and information from the latest weather forecast. The methodology combines stochastic load models and adaptive weather- load models optimally to yield an adaptive forecasting procedure especially suited for real-time control of power systems.