The Correlation of Network Topology and Power System Resilience by Using Neural Network Analysis

With the rapid increase in the connection of distributed generation and anticipated new load (electric vehicles particularly), a well-designed power system topology is critical to maintain the system's resilience against catastrophic failure. In order to understand the correlation of network topology and power system resilience, the author uses power tracing to investigate the contributions of the generations to the loads with the analysis of the network topologies. A new approach (neural networks) is then used to learn the inter-related indices between the generations and the loads under the network resilient scenarios. The results will be useful to determine the network's topology and improve power system recovery under emergencies.