Fault Identification in Distribution Network by Fusing Deep Residual Shrinkage Networks

Distribution network fault identification can increase the network's stability and safety and has great application value for smart grids. This study demonstrates a network topology based on traveling wave theory coupled with a 10kV distribution network to increase the precision and ease of fault diagnosis in the distribution network. When a fault develops in the 10 kV distribution network, this structure may assess the traveling wave propagation process, voltage, and current traveling wave signals produced, and search for a quick, precise, and reliable way to pinpoint the fault spot. Additionally, a reasonable fault location system scheme is proposed. This scheme eliminates the complex reflection process of traveling waves in the distribution network, simplifies the calculation, has a unified method, and has good algorithm universality. This system improves fault location accuracy with a maximum error of 14 meters, shortens the fault location time, greatly improves power supply reliability, and allows applying the traveling wave fault location method to the distribution network.