A Short-Term ｌｏａｄ Forecast based on TCN-BP Model

Qianzi Lu, Shiqian Li, Zhaohuan Zhu

Electrical Engineering College, Nanjing Institute of Technology, Nanjing, 210000, China

Abstract: Power system short-term ｌｏａｄ forecasting is related to the stability and safety of power system, and it also affects the formulation of power supply strategies. In reality, many factors can cause Loadchanges, so the power ｌｏａｄ has the characteristics of periodic changes and random changes. Only a certain forecasting method cannot accurately and comprehensively measure the ｌｏａｄ changes. To solve this problem, this paper proposes a short-term ｌｏａｄ forecasting method based on TCN-BP model. The actual grid ｌｏａｄ data of Zhen-jiang City is selected for simulation analysis. The experimental results show that this method can overcome the shortcomings of gradient training that based on traditional BP neural network, and can effectively improve the convergence speed and the accuracy of short-term ｌｏａｄ forecasting.

Keywords: Power system; Power ｌｏａｄ forecasting; Temporal convolutional network; BP neural network