Wavelet Packet Analysis's Application In Gear Fault Diagnosis

Resent years fault diagnosis has been paid attention to acquaint especially for great facility. Vibration signals are frequently used in mechanical fault diagnosis. However ,in many cases, because of very low signal-to-noise ratio(SNR),it is difficult to detect the fault. In this paper, a de-noising method based on Wavelet Package is introduced . The application of the method to gear fault diagnosis also gets an expected result.According to this characteristic of gearbox vibration signals, this paper presents a new vibration analysis method that is suitable for extracting fault features from non-stationary vibration signals containing weak fault information-the wavelet packet analysis method. The specific research object used in this paper is a type of locomotive hydraulic gearbox Vibration signals from gearboxes are analyzed using the method proposed and the fault feature information is obtained. The extraction of the symptom parameters is carried out using MATLAB programs.The advantaged of the method in extracting impulsive signals buried in noise is revealed by applying it to two simulated impulsive signals with additive noise.Based on the symptom information obtained by wavelet packet analysis and with the aid of the gearbox disassemble inspection results, statistical analysis method is then used to determine the criteria for the locomotive hydraulic gearbox fault diagnosis. The diagnosis results show that the symptom information from wavelet packet analysis achieved very good diagnosis results. The proposed method that combines the wavelet packet analysis and statistical analysis should be also suitable for determining the fault diagnosis criteria for other gearboxes with similar fault features.