Bus Load Prediction Method Based on SSA-GRU Neural Network

Bus load forecasting is of great significance for safe and stable dispatching of the power grid. For the characteristics of historical data of bus load changes such as complexity and time series, this paper proposes a bus load forecasting model (Sparrow Search algorithm-Gate Recurrent Unit, SSA-GRU) using the sparrow search algorithm (SSA) to optimize the parameters of gated cyclic units. The method first constructs a GRU neural network to capture the information available in the future of the time series. Secondly it uses the sparrow search algorithm to search for the optimal hyperparameters to obtain the optimal learning rate, the number of hidden layer neurons and the number of iterations, so as to improve the prediction accuracy and generalization. Finally the results validate the effectiveness and applicability of the proposed method through an example analysis of 220 kV busbars with different load attributes.