No abstract

Abstract. In this paper, short-term load forecasting of power system considering meteorological factors is studied. The power system load is divided into three parts: basic component, weather sensitive component and random component. Then the correction strategy of similar days is introduced and the meteorological factors is considered to improve the original BP model. And the similar days are determined according to the periodic characteristic of the load value and the grey relational analysis. Finally, by comparing the predicted data with the actual data, it is proved that the prediction model agrees with the actual situation and has higher prediction accuracy.