Reserch On Rolling Bearing Fault Diagnosis By Using Reduction Method Of Orthogonal Wavelet Threshold Optimization

In the mechanical equipment operation process, the rolling bearings may malfunction. The following method can be used to online monitor the malfunction. Based on orthogonal wavelet transform, use empirical mode decomposition to remove the background noise influence of the monitoring signal and leave the fault signal components. So the fault feature extraction of mechanical equipment can be guaranteed.Among all the rotating parts, rolling bearings are the most common and easily to malfunction in the real working environment. Whether the mechanical equipment can work normally is the key to the overall mechanical equipment performance. In order to avoid the l oss of equipment caused by rolling bearing fault and reduce the possibility of catastrophic accidents, online monitoring program for rolling bearings and real-time fault diagnosis was built. It has important theoretical significance and practical value for actual engineering environment.In order to obtain vibration signals of mechanical equipment under different operating conditions, test rig was used to simulate the environment of rolling bearing fault. Then, the fault type of rolling bearing can be desig ned. Test rig need to meet the following conditions: 1. The rotate speed is variable. 2. The loading is flexible. 3. The rolling bearing vibration signal can be collected fast and accurately. The experiments realized the condition monitor under three circumstances, and the corresponding vibration signals were collected as follows: 1. Normal rolling bearing; 2. Outer ring malfunctioned rolling bearing; 3. Inner ring malfunctioned rolling bearing.Based on the mathematic characteristics of orthogonal wavelet s and the optimization thresholding objective function, the collected vibration signals were denoised. On different decomposition scales, the wavelet coefficients of the the fault signals and the noisy signals have different characteristics.Besides, the hard thresholding and soft thresholding functions have diffferent advantages. By changing threshold values and the thresholding functions, the noise level was decreased while the fault signals are discriminated to the least scale. In this way, we obtain the theoretical and fundamental data for analyzing the characteristcs of the mechanical equipment malfunction.In this work, we proposed a novel thresholding choosing rule and an optimized threholding denoising method. Combined with the empirical mode decomposition characteristics extraction technics, we have applied the above mentioned method into the rolling bearings malfunction characteristics extraction. The results showed that the proposed method is efficient and applicable for rolling bearing malfunction diagnosis.