The Study Of Automobile Transmission Bearings Failure Characteristic Frequency Extraction Algorithm

Automotive Transmission Bearing Fault diagnosis can prevent damage to the gearbox,while improving vehicle safety, Aimng Automotive Transmission Bearing Fault characteristicfrequency extraction problems, based on the theory of bearing fault diagnosis, study thetransmission bearing fault frequencies extraction algorithms.Firstly, use CA2precise charge amplifier with INV306DF intelligent signal analyzer tocollect Automotive Transmission Bearing Fault vibration signal. adding random white noiseand periodic pulse noise to drown the gearbox bearing failure data, Simulate the data ofAutomotive Transmission Bearing Fault with strong noise background, provide a goodexperimental data for study.Secondly, to study the traditional gearbox bearing failure frequency extractionalgorithms, they are not able to deal with gearbox fault signal de-noising, locate the faultresonance band. According to the theory of wavelet packet associated with the delaydemodulation, then, build wavelet packet energy envelope demodulation method to make upfor the deficiencies of traditional algorithms. Wavelet packet energy envelope demodulationusing Labview to program, through the analysis of gearbox fault signal with strong noise, thealgorithm work well with gearbox failure frequency extraction.Finally, through the study of wavelet packet decomposition and reconstruction algorithm,wavelet packet energy envelope demodulation frequency selection module can be achievedthrough a digital filter. In this paper, use Dspbuilder to build wavelet packet energy envelopedemodulation frequency selection module.through the analysis of Automotive TransmissionBearing Fault data with strong noise, comparing to Labview wavelet packet energy packageenvelope demodulation results, verifies the framework of wavelet packet energy envelopedemodulation correctness frequency selection Dspbuilder module is y. Uses DspbuilderSignal Compiler to compile the selection frequency module into a VHDL language andsuccessfully compiled in Quartus, provides a feasible method for the wavelet packet energyenvelope demodulation hardware implementation.