基于改进 AlexNet 的电力系统暂态功角失稳紧急控制策略

随着新能源渗透率的提升，电网环境日益复杂，电力系统安全稳定运行也面临着新的挑战。为了满足电力系统暂态功角失稳后的实时紧急控制决策，本文采用深度学习与紧急控制相结合的方法，提出一种基于改进AlexNet 网络的电力系统暂态功角失稳紧急控制策略。首先基于改进 AlexNet 对失稳发电机功角轨迹进行预测，识别临界机群；然后定义紧急控制动作灵敏度指标，建立改进 AlexNet 灵敏度预测模型，拟合发电机功角特征与紧急控制动作灵敏度的映射关系，从而确定紧急控制的动作母线；最后以切除发电机和负荷容量最小为目标，建立紧急控制优化模型并求解最优策略，并在新英格兰 10 机 39 节点系统进行算例验证。结果表明，针对电力系统暂态功角失稳问题提出的基于深度学习的功角轨迹预测模型和紧急控制灵敏度预测模型，均有较高的预测精度。在此基础上制定的紧急控制策略能使失稳系统快速恢复稳定运行，加强电网安全稳定防御体系。

With the rising penetration of new energy resources, the power grid operation environment has grown increasingly complex, and the safe and stable operation of power systems is confronted with new challenges. To meet the requirements of real-time emergency control decision-making following transient rotor angle instability in power systems, this paper adopts a hybrid approach combining deep learning and emergency control, and proposes an emergency control strategy for transient rotor angle instability in power systems based on the improved AlexNet network. First, the improved AlexNet network is utilized to predict the rotor angle trajectories of unstable generators and identify the critical generator cluster; Next, an emergency control action sensitivity index is defined, and an improved AlexNet-based sensitivity prediction model is constructed to fit the mapping relationship between generator rotor angle characteristics and emergency control action sensitivity, thereby determining the control buses for emergency control operations; Finally, taking minimizing the amount of generator tripping and load shedding as the optimization objective, an emergency control optimization model is established to solve the optimal strategy, and case validation is conducted on the New England 10-machine 39-bus test system. The results demonstrate that both the deep learning-based rotor angle trajectory prediction model and the emergency control sensitivity prediction model proposed for the transient rotor angle instability problem of power systems achieve high prediction accuracy. The emergency control strategy formulated on this basis can enable the unstable power system to quickly restore stable operation, and enhance the power grid security and stability defense system.