基于复合潮流熵的电网连锁故障高风险场景辨识的研究

Study on High Risk Scenario Identification Method for Power System Cascading Failure Based on Hybrid Flow Entropy

电网规模的扩大化和结构的复杂化,使连锁故障对电力系统的威胁日益严重,有必要对连锁故障易发场景进行辨识。提出一种考虑系统实时运行状态和结构参数的高风险场景辨识方法,可对电网在不同运行状态下连锁故障发生概率进行评估。本方法综合考虑系统潮流分布、线路容量以及电压偏移和负荷水平,拟定复合潮流熵以反映场景连锁故障风险水平。根据复合潮流熵,提出基于花授粉算法的改进K-medoids算法对场景集进行聚类,得到指标最高的一类场景作为显著风险场景集,并建立连锁故障预测模型,计算该场景集的连锁故障概率加以验证。IEEE 30节点算例结果表明:该方法可从大量场景中有效辨识连锁故障高风险场景,可指导相关运行工作。

With the enlargement and complication of power system,the threat caused by cascading failure becomes severer.It is necessary to identify the scenario where fault occurs more possibly.This paper proposes an identification method of high risk scenario for cascading failure,considering the operation status and structure parameter.The fault probabilities of the scenarios can be assessed comprehensively in this way.Power flow distribution,line capacity,voltage deviation and load level are taken into account to establish hybrid flow entropy in this paper.So the risk level of cascading failure in each scenario can be described by the index.According to the hybrid flow entropy,the article also presents a modified K-medoids algorithm which is based on flower pollination algorithm to perform high-precision clustering on the scenario set.The cluster of highest index value is referred to as significant risk scenario set,and a cascading failure forecast model is built to calculate the fault probabilities of the foresaid scenario set.The simulation results on IEEE 30 bus system show that the proposed method can identify the high risk scenario from a mass of scenarios effectively and provide guidance for the system's operation.