Research On The Fuzzy Glustering And The Grey Theory In The Fault Diagnosid Of The Gearbox

The gear-box is the commonly used equipment, which is used to change rotating speedand transfer power. The purpose of the condition monitoring and fault diagnosis of itstransmission system is to offer scientific foundation for the pertinence maintenance, and tosave maintenance costs.The research of this subject has great economic and engineeringvalue. During the fault diagnosis for the gear-box, it is the key factors of success that pickingup the fault information effectively , judging the fault property and origin properly andselecting the effectively method of signal process and fault diagnosis .Based on the fault mechanism, vibration and structure trait of the gear-box, the theoryand method of condition monitoring and fault diagnosis for the gear-box is discussed. On thebasis of the experiment of the certain vibration test, the vibration signal of the gear-box wasobtained. At the same time, according to the vibration signal containing lots of noise, usingthe noise cancellation in wavelet process the testing signal, the problem of extracting the faintsignals from the strong background noise is solved. After picking up some feature parametersof the time-domain and frequency-domain which reflected the typical fault, more parameteridentification that can differentiate the faults in analogy methodwas carred out. Also, the faultsignal is analyzed and processed by many kinds of common method of time and frequencydomain, and the fault orientation is analyzed. At last, build on the typical fault example for thegear-box, study the pattern recognition methods based on the gear-box's state check and faultdiagnosis through the fuzzy clustering and ABO gray cognate analysis. From the diagnosticresult for the experience, it is proved that the fault diagnosis methods presented in thisdissertation are effective. And the gear-box fault can be analyzed and diagnosed validly bythese methods.