Research On Weak Signal Extraction Technology Of Gear System Failure Based On Stochastic Resonance

The application of gear system in the mechanical engineering and manyother areas are very extensive. When some parts of gear system failure, there willbe a periodic pulse force, just like wear gear, pitting gear, bending shaft, crackedshaft. They show as the spectrum characteristics of the fault signal. If we extractthe gear system fault modulation information from the weak characteristic signaland analyze them, the failure of the gear system will be apparent.Stochastic resonance system is a nonlinear system could transform part ofthe noise energy to the signal to detect the fault signal better. However, theextraction failure signal method mainly suppress the noise to detect the faultsignal, the energy of the fault signal will be suppressed too. Above all, stochasticresonance method has an important role and significance in the weak fault signaldetection. In this paper, according to the characteristics of the gear system failuresignal, using the stochastic resonance system detects failure weak signal of thegear system to achieve fault diagnosis purposes. This paper is mainly thefollowing aspects:Based on nonlinear system and the theory of stochastic resonance, discussthe impact of stochastic resonance system output by the stochastic resonancesystem parameters through modeling, and co-simulation test. through thecorresponding numerical simulation and comparison, selected a more suitablemethod In the several parameters of signal conversion method for stochasticresonance.Through the real fault vibration signals of the gear system and theoreticalmodel, this paper studied on the vibration mechanism of the early failure gearsystem and common fault signal characteristics, including the characteristics ofthe mechanism and fault spectrum of failure of gears, shafts, bearings.On the basis of the weak signal of the early failure gear system vibration,designed more focused and adaptive stochastic resonance algorithm for thefailure gear system and improved the algorithm by a large number of simulationresults and experimental analysis.The results of experimental verification by fault simulation experimentalplatform showed that the stochastic resonance algorithm given in this article canidentify the weak fault signal feature of the gear system on Machinery FaultSimulator.