Load forecasting and its calibration method considering the influence of temperature

Abstract

A calibration method for the medium/long-term load forecasting is presented. Firstly, a calibration model is built up, based on the regular pattern that the same trend is basically shared by the peak load and the maximum temperature. Secondly, a calibration method is presented. The first step is to correct the historical annual peak load and obtain the load which has no relations with the temperature. Then, forecast the peak load of the horizontal year by traditional way, regarding the corrected load obtained above as the historical data. At last, calibrate the result of forecasting according to the maximum temperature of the horizontal year and the final result is obtained. Finally, a practical case indicated that the model and method proposed by this paper can take the affection of temperature on the peak load into account in forecasting and calibrating, resulting in higher precision of load forecasting. The method has been successfully applied in an urban distribution network planning project.