In order to better mine the effective information contained in a large amount of collected data and improve the accuracy of short-term load forecasting, a short-term load forecasting method based on a hybrid model of wavelet transform and bidirectional gated recurrent unit (BiGRU) and fully-connected neural network (NN) is proposed in this paper. The wavelet transform is used to decompose the load characteristic data into high-frequency data and low-frequency data, and then, a high-frequency mixed neural network and a low-frequency mixed neural network model are built respectively to conduct prediction. In the hybrid neural network model, the load characteristic data is used as the input of the BiGRU-NN network, and the BiGRU-NN network is used to learn the load nonlinearity and time series characteristics to perform short-term load prediction. Taking the load data of Eastern Denmark as an example, the experimental results show that the proposed method has higher prediction accuracy than the GRU neural network, DNN neural network, and CNN-LSTM neural network.
