A Novel Distributed Short-Term Load Forecasting Method for Large Power Grid

C. Li, Z. Zhu, J. Shang, D. Li, and Y. Ma (PRC)


Short-term Load forecasting, large power grid, meteorological factor, distributed algorithm


Power load is quite sensitive to the meteorological factors in most of power grid. In Large Power Grid (LPG) which covers vast area, weather conditions could vary greatly from one district to the others, such as provinces in China and states in USA. It’s necessary and difficult to take meteorological factors into account in short-term load forecasting for LPG. In this paper, a novel method of short-term load forecasting is proposed for LPGs, named Subnet-summation-method (SSM). The LPG is divided into some subnets. The load of each subnet is predicted taking its weather conditions and historical data into consideration. There is difference between the load of LPG and the summation of load of LPG’s subnets. The difference is caused by different statistic-scope. Statistic scope coefficient (SSC) is defined. SSC can be predicted through analyzing the recent days. Then, the summation of the predicted load of each subnet is calculated. Finally, the summation is revised according to SSC. After revising, the predicted load of LPG is gotten. Some improvements of SSM are also proposed in this paper. According to practical application in some provinces of China, it’s proved that the new method is a simple method with high and stable accuracy.

Important Links:

Go Back