Advanced distribution automation application with mis-coordination correction algorithm based on neural network

The modern distribution grid with multiple Distributed Energy Resources (DERs) requires a distribution automation application with advanced functionality. This functionality typically includes fault localization, isolation, and restoration capabilities to achieve the maximum performance and reliability of the distribution system. In response to a fault scenario, this functionality fundamentally changes the characteristics of the distribution grid from a static circuit to a dynamic topology model. The protection algorithms must account for a dynamic topology and corresponding fault scenarios while maintaining operational control of the faulted zone(s). To address these complexities, the authors defined rules of operation which, depending on the fault scenario, will govern the system's response. A decentralized logic was designed to control individual switching devices forming a neural network with multiple nodes. The nodal connections resemble the physical connection topology between circuits and individual system devices. Paper includes system rules with a specific focus on the mis-coordination correction operations.