DC Electrocution Feature Extraction Method Based on Successive Variational Modal Decomposition

With the DC power supply system receiving widespread attention, the electrical safety of DC power supply system urgently needs in-depth research. In order to realize the fast and accurate extraction of DC electric shock features, a DC electric shock feature extraction method is proposed based on the energy entropy of successive variational mode decomposition（SVMD）. The method is based on the energy variability of animal body electric shock and non-animal body current signals, firstly, the SVMD algorithm is used to decompose the animal body electric shock and non-animal body current, and then the energy entropy of each modal component is calculated, and the energy entropy feature analysis is chosen to analyze the variability of the energy distributions of the animal electric shock and the non-animal body current in the time-frequency domain, and the energy entropy is used as a feature vector for the recognition of electric shock signals. Finally, the support vector machine（SVM） is used to realize the accurate recognition of electric shock type. The experimental test results show that the DC electric shock current features can be accurately extracted by this method, which can effectively distinguish the biological electric shock and leakage faults with high recognition accuracy, and provide a reference for the DC residual current feature extraction and recognition method.