CNN-LSTM-Based Smart Field Station Electric Load Prediction

In the smart field station, a large number of energy-using devices are deployed, and changes in power loads are complex and uncertain. To grasp the change of power load in time and make an orderly power consumption plan, this paper proposes a CNN-LSTM-based power load forecasting method for smart field stations. First, feature extraction is performed on the raw load data using CNN to capture the temporal features and local correlations of the input data. Then, the extracted feature sequences are input into an LSTM model for modeling the temporal dependence and long-term memory effects of the load data to obtain prediction results. The comparison experiments show that the scheme has good superiority and stability compared with other detection models.