Early fault diagnosis of rolling bearings is of great significance in the application of mechanical equipment, which makes the extraction of weak fault signals particularly critical by stochastic resonance (SR). Compared with bistable SR, tristable SR has stronger advantages in weak signal extraction, but the classical tristable system (CTSR) was limited by output saturation, which resulted in insufficient signal amplification ability. To solve the above problems, combined with asymmetric system whose output can be improved to a higher degree, a novel piecewise unsaturated asymmetric tristable stochastic resonance (NPUATSR) system is proposed. Through numerical simulation, it is concluded that NPUATSR output amplitude varies proportionally with the amplitude of the input, which overcomes the output saturation of CTSR. Secondly, the stationary probability density and mean first passage time of particles are derived by using adiabatic approximation theory, and the variation law caused by parameters is analyzed in combination with potential function, the internal mechanism of the system is further studied. Through the output signal-to-noise ratio (SNR), it is found that the performance advantage of NPUTASR system is the most obvious, and different parameters affect the output SNR. Finally, the adaptive genetic algorithm is used to optimize the parameters, and the proposed system is applied to early fault diagnosis on different types of bearings. After comparison with different systems, the results show that NPUATSR can effectively detect the fault frequency, and has the most outstanding advantages in spectrum amplification and anti-noise performance, which proves that NPUATSR system has significant value in practical engineering application.
Aiming at the deficiency of weak signal amplification ability in traditional bistable system and combining with the advantages of tristable system and asymmetric system, three novel of combined asymmetric tristable stochastic resonance (SR) systems are proposed by combining the classical bistable potential model with the Gaussian potential model in this paper, namely, asymmetric well depth stochastic resonance (AWDSR), asymmetric well width stochastic resonance (AWWSR) and asymmetric well depth width stochastic resonance (AWDWSR). The stochastic resonance phenomena of the above three systems under [Formula: see text] noise with the influence of parameters are analyzed, respectively. Through empirical mode decomposition and adaptive genetic algorithm to optimize the parameters for weak signal extraction, it can be found that this method can achieve stronger SR performance. After the bearing fault diagnosis of the proposed systems, the detection results of the three asymmetric systems are obviously better than the classical tristable stochastic resonance (CTSR) and symmetric stochastic resonance (SSR), and the performance advantage of AWDWSR is the greatest. This can provide some reference value for practical engineering application.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.