Weak fault detection in rotating machineries by using vibrational resonance and coupled varying-stable nonlinear systems

Abstract In the era of Industry 4.0, weak faults on a critical component must be detected timely to ensure the reliability of a critical machine. However, in general, a weak fault signal cannot be easily detected due to the presence of strong background noise and weak fault characteristics. Thus, in this work, we investigate a new weak fault detection method based on the vibrational resonance (VR) and coupled varying-stable nonlinear systems. In contrast to most existing noise cancellation fault detection methods, the proposed method can be regarded as a noise utilization method. Furthermore, in contrast to the stochastic resonance (SR) technique, which is a noise utilization fault detection method, the proposed method is easily controllable and exhibits a better performance. Moreover, resonance can be easily and adaptively achieved, and the varying-stable systems can be controlled to exhibit monostable, bistable or tristable states, by tuning the system parameters. In this paper, the nonlinear systems are coupled; therefore, the nonlinear systems are dependent on each other, and the interactions among them are extensively considered. The proposed method is validated by considering an implanted gear fault on a planetary gearbox and several naturally developed bearing faults on a fixed axis gearbox. The mechanism of noise utilization is illustrated, and different nonlinear systems with different configurations are considered. The corresponding results are compared to validate the effectiveness and advantages of the proposed method.