Cascading failure models are important for understanding the mechanism of blackouts and evaluating the control strategies to prevent the failure propagation. The evolution of cascading failure in actual power grid is a continuous dynamic process triggered by discrete events, such as initial disturbances and physical responses. In this paper, we develop an event-triggered hybrid system model to describe the dynamic process of cascading failure. In the model, the evolution of continuous states of power grid is described by differential algebraic equations and the discrete events are defined as transitions between discrete states of power grid. The model also integrates multiple physical responses including relay protection, frequency regulation and dispatching action. Based on the developed model, we propose an event-triggered simulation method of cascading failure to accelerate the simulation process. Compared with the DC power flow model, hidden failure model and topological model, the simulation results of our model are more accurate because the statistical distribution of demand loss in our model is closer to historical blackouts data. The efficiency of the proposed event-triggered method is demonstrated by comparing our model with the time-driven model and three existing models. The experimental results show that our model can trade off the simulation accuracy and time consumption. Note to Practitioners—This paper focuses on modeling the dynamic process of cascading failure with multiple physical responses in power grid. We develop an event-triggered hybrid system model for cascading failure. In the model, the continuous dynamics of power grid and discrete events triggering the evolution of cascading failure are all described by the framework of hybrid system, which is a good example of modeling the hybrid system for automation researchers and engineers. By this way, the model is more accurate in describing the actual characteristics of cascading failure in power grid, and thus supporting the design and evaluation of control strategies for improving the stability of power grid. Based on the developed model, we propose an event-triggered simulation method of cascading failure, which aims to improve simulation accuracy while potentially reducing time consumption. In practice, the model can make fast control strategies to prevent the failure propagation.

中文翻译：

---

电网连锁故障的事件触发混合系统模型

级联故障模型对于理解停电机制和评估控制策略以防止故障传播非常重要。实际电网中连锁故障的演化是一个由初始扰动和物理响应等离散事件触发的连续动态过程。在本文中，我们开发了一个事件触发的混合系统模型来描述级联故障的动态过程。在模型中，电网连续状态的演化由微分代数方程描述，离散事件定义为电网离散状态之间的过渡。该模型还集成了多种物理响应，包括继电保护、频率调节和调度动作。基于开发的模型，我们提出了一种级联故障的事件触发仿真方法来加速仿真过程。与直流潮流模型、隐性故障模型和拓扑模型相比，我们模型的仿真结果更加准确，因为我们模型中需求损失的统计分布更接近历史停电数据。通过将我们的模型与时间驱动模型和三个现有模型进行比较，证明了所提出的事件触发方法的效率。实验结果表明，我们的模型可以兼顾仿真精度和时间消耗。我们模型的模拟结果更准确，因为我们模型中需求损失的统计分布更接近历史停电数据。通过将我们的模型与时间驱动模型和三个现有模型进行比较，证明了所提出的事件触发方法的效率。实验结果表明，我们的模型可以兼顾仿真精度和时间消耗。我们模型的模拟结果更准确，因为我们模型中需求损失的统计分布更接近历史停电数据。通过将我们的模型与时间驱动模型和三个现有模型进行比较，证明了所提出的事件触发方法的效率。实验结果表明，我们的模型可以兼顾仿真精度和时间消耗。从业者须知——本文侧重于对电网中具有多个物理响应的级联故障的动态过程进行建模。我们为级联故障开发了一个事件触发的混合系统模型。在模型中，电网的连续动态和触发级联故障演化的离散事件都用混合系统的框架来描述，是自动化研究人员和工程师对混合系统进行建模的一个很好的例子。通过这种方式，该模型更准确地描述了电网连锁故障的实际特征，从而为提高电网稳定性的控制策略的设计和评估提供支持。基于所开发的模型，我们提出了一种事件触发的级联故障仿真方法，其目的是提高模拟精度，同时可能减少时间消耗。在实践中，该模型可以制定快速控制策略来防止故障传播。