Research on risk assessment algorithm and experimental investigation for power systems under extreme weather conditions

To characterize the spatiotemporal evolution and superposition effects of hazardous weather on power system outage risk, a mathematical-modeling-based spatiotemporal risk assessment algorithm for power systems under hazardous weather conditions is proposed and evaluated through simulation experiments. First, intensity models are established for three typical hazardous weather scenarios—storms represented by typhoons, frost, and rainstorms—by incorporating their movement trajectories and impact ranges. Subsequently, two compound hazardous weather scenarios, namely “storm + rainstorm” and “storm + frost,” are constructed. The failure probability of each power system unit is calculated according to the hazardous weather intensity, and its spatiotemporal evolution characteristics are analyzed. Finally, low-, medium-, and high-risk regions are classified based on the spatiotemporal distribution of failure probabilities. The results show that the failure probability of distribution equipment exhibits pronounced spatiotemporal migration as the centers of hazardous weather events move, with risk regions dynamically evolving along the movement paths of the hazardous weather systems. In the storm + rainstorm scenario, the medium- and high-risk regions cover a wider area and exhibit a more pronounced risk superposition effect. In contrast, under the storm + frost scenario, high-risk regions are mainly concentrated in the overlapping areas affected by both the storm and frost. The proposed algorithm can effectively track the dynamic migration paths and evolution trends of high-risk regions for power equipment failures under compound hazardous weather conditions and generate intuitive spatiotemporal risk maps, thereby providing technical support for optimizing emergency resource allocation in power systems before hazardous weather events and for implementing precise early warning and proactive control during such events..