摘要：To understand the systemic propagation and interdependencies of various risk factors in large-scale power outages, a novel risk evolution framework is proposed, incorporating Event Tree Analysis (ETA), directedweighted Complex Network (CN), and risk chain mining. A network comprising 37 nodes and 130 edges is constructed using data from 29 large-scale power outages spanning from 1965 to 2025. Topological analysis identifies critical propagation nodes, including power plant failures and grid instability. Robustness tests reveal the network's resilience to random disruptions but also expose its vulnerability to targeted attacks. Furthermore, high-risk causal chains are directly extracted and classified from the network, and path mining further uncovers 24 high-risk six-node chains, which are categorized into four distinct propagation patterns. This integrated framework offers a data-driven methodology for characterizing large-scale power outage risks, yielding structurally suggestive priorities to inform prevention and mitigation strategies for power systems.

关键词：Large-scale power outage；Risk propagation；Complex network；Risk chain

文献来源：ELECTRIC POWER SYSTEMS RESEARCH

WOS链接：https://webofscience.clarivate.cn/wos/alldb/full-record/WOS:001886096500001