Algorithm for extracting spectral features of substation high-voltage isolation switch card shibboleth faults

Abstract

Abstract Disconnectors are susceptible to equipment defects and hidden dangers due to the harsh natural environment, the impact of load fluctuations in the power grid, and manufacturers’ manufacturing processes. The current production of commonly used fault diagnosis technology is limited, and the number of features extracted from on-site data cannot timely and accurately reflect the equipment failure conditions. Therefore, this paper uses the substation high-voltage isolation switch jamming fault spectrum feature extraction algorithm, the use of high-frequency acquisition technology to establish the power characteristics of mathematical models and current characteristics of mathematical models, and the normal operation and jamming fault data for classification. To establish a sample library of fault models for the current time curve of the knife gate action process, for the typical faults of the disconnecting switch (not in place, reversal and jams, etc.), simulation is carried out in the laboratory environment to form a series of sample libraries of the motor fault waveforms; it is shown by the test that it is feasible to utilize the power of the motor of the high-voltage disconnecting switch during the opening and closing of the gate action process to determine the faults of the high-voltage disconnecting switch mechanism.