Regional partition based rapid prediction of power outages in distributed systems following flooding

• 1.The siting flexibility of power distribution equipment is weak, which is prone to outage failures caused by flooding and inundation from rainstorms. • 2. 2D hydrodynamic model can realize rainstorm inundation simulation for power outage prediction, but there is a time bottleneck. • 3. A fast prediction method for distribution network inundation outage based on area segmentation is proposed to realize real-time prediction of rainstorm outage. The inappropriate siting of power distribution equipment can make these facilities susceptible to flooding, triggering power outages and seriously impacting the reliability of power supplies. A two-dimensional hydrodynamic model can be used to simulate flood inundation that can be used in outage prediction. However, the required computational time is proportional to the square of the simulation area, making it difficult to apply for online outage prediction during large-scale flood events. Therefore, a fast prediction method for flood inundation outage in distribution network based on area partition is proposed. Firstly, a two-dimensional hydrodynamic model is established based on the geographic information data related to a once-in-50-years extreme rainfall event, and then an inundation simulation is undertaken over a complete urban area to determine the high-risk area for inundation outage. Then, the watershed system is extracted, and the high-risk area for inundation outage in the catchment area is initially selected and identified as the core area of inundation. Next, an optimization model is established, targeting the optimized simulation time, with the constraints of the simulation deviation between the core area and the complete area. A genetic algorithm is then used to solve the size of the optimal simulation area, and ultimately determine the optimal simulation area. Finally, the obtained optimal simulation area is used to simulate flood inundation to provide a rapid prediction of the distribution network power outage caused by the rainstorm and consequential flooding. Simulation analysis demonstrates that the proposed method balances the accuracy and speed of inundation simulation, thereby providing crucial support for grid emergency management.