A new method of fault diagnosis for high-voltage circuit-breakers based on Hilbert-Huang Transform

In order to improve the performance of conventional methods of fault diagnosis for "high-voltage circuit-breakers (HVCBs)", "Hilbert-Huang Transform (HHT)" is introduced to analyze mechanical vibration signals. HHT is composed of "empirical mode decomposition (EMD)" and "Hilbert spectrum analysis". EMD method is adaptive to the analysis scale, with which complicated vibration signals can be decomposed into a finite number of "intrinsic mode functions (IMFs)" that admit well-behaved Hilbert transforms. The Hilbert spectrum analysis can present us the accurate time-frequency distribution of vibration signals with high resolution. Based on HHT, entropy vectors are extracted as features from Hilbert spectrum, which represent the evenness of energy distributions in different frequency bands. Such features of normal and abnormal vibration signals are used to train a two-layer feed-forward backpropagation neural network, with which different patterns of the signals can be recognized determinately.