Fault Location in Distribution Networks Based on Voltage Sag Monitoring Data

With the rapid increase in power quality monitoring data and the maturity of data analysis and processing technologies, integrating power quality monitoring data with actual production needs to explore the application value of this data has become an important research focus in the field of power quality. This paper proposes a method for fault location based on voltage sag monitoring data, utilizing a BP neural network to model the mapping relationship between residual voltage during voltage sags and fault locations. First, a batch simulation of the fault data set is conducted to obtain voltage sag data, followed by offline training of the BP neural network to develop a neural network model for fault location. Finally, the online monitored voltage sag data is input into the pre-trained neural network for online application, resulting in the identification of the fault location. This method converts the output results of the BP neural network into fault location determination results, establishing a fault location model composed of the trained neural network and the fault location determination program, which can quickly predict the location of faults.