High-Impedance Fault Identification Using Cyclostationary Characteristic Analysis

•The High-Impedance Fault can cause deaths of people and animals.•The cyclostationary analysis extracts information from the signals.•The noise has a minimized influence in the cyclostationary characteristic analysis.•The main difficulty found by overcurrent protection the low fault overcurrent level. Conventional overcurrent-based protection systems are generally not sensitive to high impedance faults (HIFs) since they have a low overcurrent amplitude. This type of fault causes damages to dealers and can provoke the deaths of people and animals. Therefore, different methods for identifying HIFs in electric power distribution systems have been proposed. However, besides the low fault overcurrent level be still a problem, the noise interference on signals is also a difficulty for non-conventional methods. Therefore, this work proposes a reliable method based on statistical characteristics for identifying HIFs regardless of the noise interference and the overcurrent level. Specifically, the method uses cyclostationary characteristic analysis to extract cyclic autocorrelation information from the signals of interest by calculating the cyclic spectral density function. From this information, HIFs can be properly identified among other power system disturbances. The performance of the method was assessed with actual HIF data and with realistic HIF simulations, presenting promising results.