A feature extraction and pattern recognition method of vibration signals of high voltage distribution equipment based on LabVIEW

Focusing on the non-stationary characteristics of mechanical vibration signals of high voltage distribution equipment, a novel feature extraction method of vibration signals was proposed based on a joint analysis of wavelet packet decomposition and reconstruction, Hilbert transform and normalized energy spectrum which all can be performed on the platform of LabVIEW. The signal was decomposed into finer bands by wavelet packet transform, then the signal envelope was extracted using Hilbert transform and the normalized energy spectrum can be got by calculating the energy of the envelope, the spectrum can be regarded as an eigen feature quantity of the distribution equipment diagnostic system. Simulation on the distribution equipment was conducted and vibration signals were collected in three different conditions of which are the normal condition, opening spring loose and friction in solenoid. The vibration signals were analyzed based on the characteristics of time-frequency and K-nearest neighbor algorithm for different conditions eigen quantity pattern recognition can be used to make a detailed analysis of the signals. Experiment results show that each element of the normalized energy spectrum vector of the normal signal is evenly distributed; while the elements of fault signal normalized energy spectrum vectors are remarkably varied. The 93.3% accuracy and fast recognition speed verify the feasibility of the mechanical fault diagnosis approach for high voltage distribution equipment.