Rotating Machinery Fault Feature Extraction And Analysis Techniques

In modern industrial production, rotating machinery has always been one of the important equipment in the rotating chain, therefore, it became the essential work in the industrial production to foresee and diagnosis the rotating machinery fault. In recent years, rotating machinery fault diagnosis technique is one of new technologies with more rapid development, while the most important technology of fault diagnosis is extract and analyze fault information.How to extract and analyze fault characteristics effectively is the focuses issues in current study of fault diagnosis technology.This paper make the common faults of rotating machinery as study object, analyze the failure mechanism of the common mechanical failures, and apply the fault feature extraction and analysis methods for extract and analyze fault information, and finally to distinguish the source of the failure. The main content is as follows:1.Analyze common failure mechanism for rotating machinery, describe the characteristics and manifestations in detail of the bearing failure, gear failure and rotor misalignment fault. Two different design of the vibration signal acquisition system were realized by equal time sampling and equal angle sampling.2.Focuses on describe the feature extraction and analysis methods based on the signal of rotating machinery fault and introduce the correlation analysis of amplitude domain analysis, time domain analysis, frequency domain analysis and order analysis, and has done a simulation example.3.Extract, analyze and diagnose the typical mechanical fault that set up by fault diagnosis test bench. Bearing failure, gear failure and rotor misalignment fault in different conditions were simulated by test bench. Use hilbert transform and order analysis to extract signal features based on equal time sampling and equal angle sampling. Finally, analyze the fault sources accordance with the failure mechanism.Proved by experimental analysis:Hilbert demodulation spectral analysis and order analysis could accurately extract the typical mechanical failure characteristics under different conditions, and achieve the correct identification of the fault source.