부산대학교 도서관

The Spring Festival is an important festival for family reunion, when the crowd returns to their hometown together, causing transfer and concentration of the power load. Accurate load forecasting can effectively enhance reliability of power supply in the Spring Festival. In order to solve the problem of load forecasting during the Spring Festival in Sichuan Province, the t-test is used to analyze the change trend of load during the Spring Festival. Then, considering the randomness and non-linear relationship of power load, the predecessor model is used in other fields, the accuracy of the power load during the Spring Festival in Sichuan Province. It is indicated in the result that there are laws for change in load during this period, and that Feedforward Neural Network has good effect on forecasting of load, providing bases for power load dispatch.