Research on Power ｌｏａｄ  Forecasting based on Time Series 

Shiqian Li, Jiaxin Li, Xingjian Shi 

School of Power Engineering, Nanjing University of Engineering, Nanjing, 211167, China 

Abstract: At present, with the widespread popularity of power facilities and the rapid development of the power grid, a set of research models based on time series improvement has been introduced for the problem of power system ｌｏａｄ forecasting. This method is based on deep mining analysis of the regularity of the power ｌｏａｄ  itself, mathematically processes the power ｌｏａｄ forecast in actual operation, draws up parameters based on correlation functions, and transforms into a stable time series for forecasting. The test results show that in the short-term ｌｏａｄ forecasting, the forecasting result is in good agreement with the actual value, the error is less than 2%, the data is accurate, and it has high reliability and measurement value.

Keywords: Power system; ｌｏａｄ forecast; Function formulation parameters; Stationary time series