Forecasting residential air conditioning loads

Abstract

A doubly censored Tobit model is used to forecast hourly air-conditioner usage for individual households. The model worked well over a wide range of temperatures, 9–38°C, making it possible to accurately forecast the electricity load for a variety of demand response applications including operational reserves for renewable energy integration. Individual models are simulated and summed to obtain aggregate forecasts and confidence intervals. The model allows for correlation between the individual shocks that occur in a region. This approach gives substantially more accurate results than the moving average method typically used for forecasting and measuring direct load control. Applying the model to data from three U.S. utilities produced mean square error values from 0.027 to 0.041 with average load values per customer ranging from 0.49 to 0.62kW.