Order Analysis Method Based On The Fault Characteristic Frequency And Its Application To Roller Bearing Fault Diagnosis

Roller bearing has many advantages such as high efficiency, small frictional resistance, convenient assembly and easy lubrication, so it is applied widely in the rotary machines and is the core component of the mechanical equipments, especially the rotary machines. Harsh work condition and complex load lead to a higher failure rate of roller bearings. The research on condition monitoring and fault diagnosis technology of roller bearings is of essential importance to the safety and stability of mechanical equipments.Condition monitoring and fault diagnosis of roller bearings has two features:First, the monitoring signals measured in a roller bearing under variable speed and load mode are usually non-stationary signals; Second, there are many modulation frequencies in the fault signals and it is difficult to extract and recognize them. This thesis, funded by the project "Sparse Signal Decomposition Based on Multi-scale Chirplet and Its Application to Mechanical Fault Diagnosis "(Project’s Serial Number:50875078) supported by National Natural Science Foundation of China, and the project "Research on condition monitoring and fault diagnosis of large wind turbine"(Project’s Serial Number:20090161110006) supported by The National High Technology Research and Development of China863Program, researched the fault diagnosis problems of roller bearings.The main research works include:(1) The developments of the fault diagnosis technologies of rolling bearings and the main diagnosis methods at present were introduced. The methods of order tracking were presented and analyzed.(2) According to the structural features and working principle of roller bearings, the fault characteristic frequency formulas of different parts have been derived. Envelope demodulation, sparse signal decomposition based on multi-scale chirplet and order analysis were introduced briefly. This thesis fuses these methods into a new method:Order Analysis Based on the Fault Characteristic Frequency, the advantages of the new method were presented.(3) Aiming at the feature that the envelope of vibration signals of the roller bearing with variable rotating speed contains a lot of fault information, a method of envelope order spectrum based on the fault characteristic frequency is proposed. In the proposed method, the fault characteristic frequency is extracted from the envelope of vibration signal using the method of sparse signal decomposition based on multi-scale chirplet, and then the order analysis based on the fault characteristic frequency on the envelope signal is carried out. The method is applied to the fault diagnosis of the roller bearing with variable rotating speed. The result shows that the proposed method is effective.(4) Aiming at the feature that the envelope of vibration signals of the roller bearing with variable rotating speed contains a cluster of harmonics of the fault characteristic frequency, a method of order analysis based on the fault characteristic frequency on component signals of roller bearings is proposed. In the proposed method, the fault characteristic frequency and its main harmonics are extracted from the envelope of vibration signals using the method of sparse signal decomposition based on multi-scale chirplet, and then the order analysis based on the fault characteristic frequency on the sum of component signals is carried out. This method is applied to the fault diagnosis of the roller bearing with variable rotating speed. The result shows that the proposed method has strong noise immunity and is good at identifying the component signals, and is very effective.(5) Aiming at the feature that the vibration signals always take on apparent nonlinearity, non-Guassianness and nonstationarity when the roller bearing fails, a method of order bispectrum analysis based on the fault characteristic frequency is proposed. In the proposed method, the fault characteristic frequency of the roller bearing is extracted from the envelope signal using the method of sparse signal decomposition based on multi-scale chirplet. Based on the fault characteristic frequency, the original non-stationary signal is transformed into stationary signal, and then the bispectral analysis is carried out. Simulation and practical application examples verify that the proposed method is very effective in extracting the fault characteristic frequency of roller bearing and highlighting the fault features.In this thesis, according to the structural and fault signal characteristics of roller bearings, the method of Order Analysis Based on the Fault Characteristic Frequency is proposed. Based on the method, the thesis proposes a method of envelope order spectrum, a method of order analysis on component signals and an order bispectrum method. Simulation and practical application examples verify that these methods can be effectively applied to the fault diagnosis of roller bearings under time-varying rotational speed condition.