Power load prediction method and terminal

The invention is suitable for the technical field of power, and provides a power load prediction method and a terminal, and the method comprises the steps: obtaining historical load data and influence factor information of a power grid on a reference day, wherein the reference day is a day before a prediction day, and the influence factor information comprises air temperature information and holiday and festival information; obtaining influence factor information of the prediction day; according to the influence factor information of the reference day and the influence factor information of the prediction day, adjusting historical load data of the power grid on the reference day; and inputting the adjusted historical load data of the power grid on the reference day into a pre-trained neural network prediction model for prediction to obtain load data on the prediction day. The prediction accuracy of the neural network prediction model can be effectively improved. 本发明适用于电力技术领域，提供了一种电力负荷预测方法及终端，包括：获取参考日电网的历史负荷数据