Research On Method Of Hydraulic System's Fault Feature Extraction & Fault Classification Based On Wavelet Theory

There are a large number of non-stationary signals in the fields of condition monitoring and fault diagnosis for hydraulic equipment. Researching and developing effective engineering methods for processing non-stationary signals are necessary for promoting sustained development of fault diagnosis technology. Over the last few years, rapid expanding methods and theories for processing non-stationary signals, especially wavelet theory, provided powerful tools for condition monitoring and fault diagnosis for hydraulic equipment. In this paper, application problems of wavelet theory are investigated, such as signal de-noising, fault feature extraction, fault classification and signal demodulation analysis. The main research works are as follows:The paper first introduces the actuality & significance of hydraulic system's fault diagnose, and the theory and characteristic of wavelet and wavelet packet, the application trends of wavelet analysis in signal filtering; indicate the insufficiency of threshold denoising, fusion the strongpoint of adaptive filtering, and then bring forward integratingwavelet transform & adaptive filtering------an adaptive filtering method based on wavelettransform. When hydraulic power system running in different state, the energy of current signal wide change in frequency domain and contain abundance fault information, combine wavelet packet can decompose high frequency and low frequency simultaneity, this paper proposes a new technique of multi-band energy feature extraction and the detail feature extraction of its main band for current signal.The method is used to the noise reduction of the pressure signal measured in hydraulic equipment. The result indicates that the method can eliminate the noise in the signal more effectively than the usual threshold filter. Combine wavelet packet energy feature with ART1 neural network, and then present a fault classification method based on the combination of wavelet packet analysis and ART1 neural network. The fault instance of the current signal on hydraulic power system proves this method can distinguish 8 genus faults and is very effective and feasible. This text compared the principle and advantage& disadvantage of some common signal envelope analysis method, using Morlet wavelet distill signal's envelope curve. The text aim at envelope curve's time domain feature parameter, combine with gray fault diagnose mode to diagnose hydraulic system. The experiment result indicates: using this method to diagnose the fault in hydraulic power system is effective, it is necessary to develop the research of combination wavelet transform with some other analysis method and empolder practicality technology.