Capacitance status identification of subway vehicles based on optimized VMD and energy relative entropy

Aiming at the problem that there is no obvious symptom of capacitance performance degradation on subway vehicles, a capacitance status identification method based on optimized variational mode decomposition (VMD) and energy relative entropy was proposed. By establishing a Matlab circuit model, the output voltage signals of the capacitance at normal status or different degradation conditions were extracted, then the characteristic samples were obtained by decomposition of optimized VMD. And the relative entropy analysis of the energy eigenvectors of the eigenmode components at the above status was carried out to obtain the identification threshold of capacitance degradation. In practical application, the relative entropy value of the energy of the circuit under test and the normal status was compared with the identification threshold to complete the capacitance status identification. The analysis result shows that this method can identify the capacitance status simply and effectively, and the accuracy is 93.3%.