| The diagnosis of mechanical faults in urban rail traction systems has a significant importance on both safety and reliability, which can avoid train crashes. In order to extract the fault features effectively, local characteristic-scale decomposition (LCD) is used in vibration signal analysis of urban railcar bearings. The contributions and conclusions are made as follows:In the light of the characteristics of urban railcar bearing vibration signal, the relationship between the non-stationary, non-Gaussian properties and bearing faults are analyzed. As a result, a framework for bearing fault detection by means of adaptive signal decomposition presented. Furthermore, characteristics and shortcomings of the conventional adaptive signal decomposition methods (experience mode decomposition, local mean decomposition and local characteristic-scale decomposition) are demonstrated.The local characteristic-scale decomposition (LCD) can decompose any complicated signal into a number of intrinsic scale component whose instantaneous frequencies own physical sense and a residual. This characteristic makes LCD especially suitable for processing the non-stationary signal. The local characteristic-scale decomposition (LCD) is compared with empirical mode decomposition (EMD) as well by analyzing the computer-generated signal. The results indicate that the LCD is superior to the EMD method in retraining the end effect, iteration time and the accuracy of the instantaneous characteristics. These advantages illustrate that LCD is better suited to vibration signal on-line analysis.By summing up the results of the above study, fault Diagnosis System of Rolling Bearings based on Ethernet is developed, which can diagnosis and warn the fault of bearing of rail vehicle early by monitoring temperature and vibration signals via Ethernet. Commissioning show that:the system had very strong robustness to figure out all fault types of bearing of rail vehicle effectively. |
PDF Full Text Request