Early fault diagnosis method for gearbox based on second-order underdamped SR and its application

Stochastic resonance (SR) has been proved to be an effective method to extract weak-fault signals from gearboxes under strong noise. In this paper, we propose a method based on underdamped second-order SR with a new potential function. In our testing, we added an appropriate amount of Gaussian white noise to the original signal. When it matched the periodic signal and the potential function, the method showed good noise reduction ability, highlighted the fault signal and extracted fault information more effectively than the traditional SR method. The ant colony algorithm was used to optimize the potential parameters of the SR potential model and obtain the best output signal-to-noise ratio (SNR). We verified the proposed method by computer simulation and experimental verification, demonstrating that it produced better output than that of the traditional SR method.