电动汽车充电负荷预测对保障电网稳定运行至关重要。针对现有方法忽视多因素影响且难以兼顾长短期依赖的问题，本文提出一种结合时域卷积网络（Temporal Convolutional Networks，TCN）与注意力机制的长短时记忆神经网络（Attention Long Short-Term Memory，AttLSTM）的多特征输入充电负荷预测模型。3 个公开数据集的试验结果表明，该模型预测准确性优于对比模型，平均绝对误差（Mean Absolute Error，MAE）和均方误差（Mean Squared Error，MSE）相较于最优基线模型分别降低 4.7% 和 12.6%，从而为电网稳定运行和调度优化提供支撑。

Electric vehicle charging load forecasting is crucial for ensuring the stable operation of the power grid.In view of the problems that the existing methods ignore the influence of multiple factors and have difficulty taking into account both long-term and short-term dependencies, this paper proposes a multi-feature input charging load prediction model combining Temporal Convolutional Networks (TCN) and Attention-enhanced Long Short-Term Memory (AttLSTM) neural network. The experimental results of three public datasets show that the prediction accuracy of this model is superior to that of the comparison model. The Mean Absolute Error (MAE) and Mean Squared Error (MSE) are reduced by 4.7% and 12.6% respectively compared with the optimal baseline model. This model can provide support for the stable operation of the power grid and the optimization of dispatching.