Research Of Weak Signal Enhancement Method Based On Vibrational Resonance And Logical Stochastic Resonance

Generally speaking,the vibration signals which acquired from the mechanical equipment are always overwhelmed by heavy background noise.These heavy background noises will influence the diagnosis results,which is not beneficial to detecting and diagnosing the machinery effectively.Weak signal amplification belongs to the category of signal processing,and its essence is that how to effectively extract the weak feature information which buried in heavy background noise.Therefore,how to effectively extract the feature information of the weak signal has been one of the important contents of machinery fault diagnosis.In this thesis,a physical model based on the enhancement mechanism of high-frequency sinusoidal periodic signal is studied,which is called vibration resonance(VR).The difference between the enhancement mechanism of YR and stochastic resonance(SR)lies in:SR model uses noise to enhance weak signal,while VR amplifies the weak signal by using the disturbance of high frequency signal in the potential well.The related engineering cases show that the physical model has good fault detection capability.In recent years,some scholars abroad with the help of the traditional SR physical model,and they proposed a square wave signal enhancement mechanism that can be used to the logical detection,named as logic SR(LSR).Murali and other scholars utilize the LSR theoretical background to construct the circuit model.Their simulation results show that the required logical behaviors can be obtained by establishing the circuit model.In this thesis,tristable potential is combined with the traditional LSR physical model to achieve the reliable logical behavior through the parameter adjustment method.Resonance characteristic curve shows that the physical model has LSR phenomenon.By searching for the optimal system parameters,'AND' and 'OR'logical operations are verified.The analysis results of simulation signal show that the logical output waveform quality of tristable LSR is better than that of the bista'ble one.The results of quantitative analyses show that the waveform quality of the tristable output signal is better than that of the bistable one.This study shows that the optimal logical output can be obtained by optimizing the system parameters.To sum up,this thesis mainly focuses on the research of the weak signal enhancement method.First of all,this thesis studied the weak signal detection model based on VR and its application in bearing fault detection.Compared with previous research results,this research method shows better fault detection ability,with higher signal-to-noise Ratio(SNR),which is beneficial to the feature extraction of rolling bearings.Secondly,the physical model of LSR is preliminarily studied.In this physical model,the influence of different system parameters on the detection precision of weak square wave is analyzed.