Analysis of Weak Signal Detection Based on Tri-Stable System under Poisson White Noise

Abstract

Poisson white noise is a typical non-Gaussian noise that can induce stochastic resonance to detect feature signals submerged within it. In this paper, we propose a detection method for overdamped tri-stable stochastic resonance in Poisson white noise environment. Based on the model of over-damped tri-stable stochastic resonance. To begin with, the influence of the parameters of the tri-stable system on the potential function is studied. Then, the cross-correlation coefficient C_sx is used to measure Poisson white noise under different pulse arrival rates β, system structural parameters a, b, c, and noise intensity D the law of action on system resonance. Ultimately, the system is successfully applied to the fault diagnosis of bearing under variable speed conditions and the background of Poisson white noise, and the simulation and experimental results are compared with the classical bistable system. The findings suggest that the performance of the system significantly surpasses that of the classical bistable system.