A Novel Forecasting Method for Short-term Load based on TCN-GRU Model
Xiaoyan Hu, Bingjie Li, Jing Shi, Hua Li, Liu Guojing
Abstract
In order to improve the accuracy of short-term electric load forecasting and provide stronger assurance for the stable operation of the electric power system, a short-term load forecasting method, TCN-GRU, which combines time convolutional network (TCN) and gated recurrent unit (GRU) is proposed in this paper. This method comprehensively considers the timing characteristics and non-timing characteristics of the data. The short-term electric load prediction is realized by the TCN model to realize the further feature extraction of the time series features and the nonlinear fitting ability of the GRU model. Based on the electric load data of an industry in Nanjing, Jiangsu Province, the load forecasting ability of the TCN-GRU model is verified. Experiments show that the proposed method has a great advantage over the other deep learning methods.