Research On Feature Extraction And Identification Of Shaft Orbit For Rotating Machinery

With the rapid development of society, the demand for large rotating machine is increasing, at the same time the fault diagnosis technology of rotating machine has been developed rapidly. As is known to all, the main research direction of fault diagnosis for rotating machine at home and abroad is the three aspects of failure mechanism, feature extraction and axis trajectory identification. As an important representation of the rotating machine, theaxis trajectory is the core of the fault diagnosis of rotating machine.The main work of this paper is to study the connection between the axis orbit pattern and the fault by the feature extraction and identification of the axis orbit. First, several common fault mechanism and vibration characteristics are analyzed and summarized, in order to lay the foundation for the identification and classification of fault, then through the establishment of mathematical model and numerical simulation to carry on the fault simulation,and make a comparative analysis of the connection between the axis trajectory graphics and fault. In the feature extraction of axis orbit combined two-dimensional shape invariant moment method with the Fourier descriptor method, and then use the information fusion and the decision-making function of the D-S evidence theory to identify graphics of axis orbits and compared results with that of traditional Euclidean distance metric method, showed a significant advantage. Finally several faults were simulated on the rotating mechanism of control and fault diagnosis test in the laboratory, and summarized the connection between the axis locus graphics and the fault by analyzing the experimental results.