为应对分布式可再生能源规模化接入乡村配电网导致的消纳难题，提出了一种数据-知识混合驱动的乡村配电网电氢协同调度策略。首先，基于XGBoost提取关键特征，并输入BiLSTM获取氢燃料汽车加氢需求预测序列，为配电网电氢协同调度提供精确数据边界。然后，设计了电力-交通网络耦合的电氢协同双层滚动优化架构，上层建立考虑氢气运输延时效应的能源集线器模型，制定配电网最优调度策略，并下发站间氢气互济量；下层建立基于宏观交通流的动态路径规划模型，依据县乡道路交通状态动态更新最优运氢路线，并实时反馈运氢延时信息至上层，触发配电网调度滚动修正。研究表明，所提策略能够有效提升乡村配电网可再生能源消纳率。

To address the renewable energy accommodation challenge caused by the large-scale integration of distributed renewable energy into rural distribution networks, a data-knowledge hybrid-driven electricity-hydrogen coordinated dispatch strategy is proposed. Firstly, XGBoost is used to extract key features, which are then input into a BiLSTM model to forecast the hydrogen refueling demand of hydrogen fuel cell vehicles, providing accurate data boundaries for coordinated dispatch. Then, a bi-level rolling optimization framework coupling the power and transportation networks is developed. The upper level establishes an energy hub model considering hydrogen transportation delays to determine the optimal dispatch strategy of the distribution network and the inter-station hydrogen mutual-aid quantities. The lower level constructs a dynamic route planning model based on macroscopic traffic flow to update the optimal hydrogen delivery routes according to county-township road traffic states, and feeds the transportation delay information back to the upper level in real time to trigger rolling dispatch revisions. Comparative results show that the proposed strategy effectively improves the renewable energy accommodation rate of rural distribution networks.