Weak Signal Detection And Recovery Based On The Novel Potential Well Model Of Stochastic Resonance

The weak signal detection method based on stochastic resonance is to use the noise energy transfer mechanism to extract and identify the weak signal submerged in the noise.At present,the stochastic resonance weak signal detection theory based on the classic bistable model has become more mature,and it has been found that the classic bistable model has problems of output saturation and high barrier.Therefore,in order to further enhance the enhanced detection effect of the stochastic resonance system,a novel potential well model is constructed based on the piecewise linearized bistable model and the applied signal stochastic resonance model,and research will be carried out around the novel potential well model.Firstly,the novel potential well model is theoretically analyzed to derive formulas of the Kramers escape rate and the output signal-to-noise ratio.The detection performance of the stochastic resonance system based on the novel potential well model and the classic bistable system driven by periodic signal and impact signal are analyzed.The research results show that the stochastic resonance system based on the novel potential well model has better detection effect.The output of the stochastic resonance system will produce waveform distortion,and the reason for the distortion of the output signal waveform of the stochastic resonance system is analyzed based on the principle of combining signal processing and particle dynamics.The signal recovery system was established to recover the stochastic resonance output signal,so as to realize the recognition of the signal waveform submerged in the noise environment.Finally,in order to verify the effectiveness of the research method in practical problems,the rolling bearing is taken as the research object.The stochastic resonance system based on the novel potential well model is used to detect the inner race,outer race and bearing rolling element fault signals of the bearing drive end,and compared with the detection results of the classic bistable system.The results show that the output of the stochastic resonance system based on the novel potential well model has a better detection effect.The recovery system is used to realize the waveform recovery of the output signal of the stochastic resonance system to verify the theory and method studied.The research results based on the novel potential well model further enriches the stochastic resonance theory and provides more possibilities for the detection of weak characteristic signals in the field of practical engineering.