Applications of stochastic resonance to machinery fault detection: A review and tutorial

Abstract

Abstract Fault detection is a key tool to ensure the safety and reliability of machinery. In machinery fault detection, signal processing methods are extensively applied to extract fault characteristics. Widely used signal processing methods attempt to eliminate the noise imbedded in signals for discovering fault characteristics. Different from widely used signal processing methods, stochastic resonance (SR) is able to utilize the noise imbedded in signals to extract weak fault characteristics from the signals. Therefore, it has been extensively applied to fault characteristic extraction and machinery fault detection. Up to now, massive literature on the applications of SR to machinery fault detection has been published in academic journals, conference proceedings, etc. This paper attempts to survey and summarize the current progress of SR applied in machinery fault detection, providing comprehensive references for researchers concerning with the subject and further helping them identify future trends for research. First, this paper elaborates SR from its original mechanism to fundamental theory. Then, the literature on machinery fault detection using SR is reviewed in terms of critical rotary components prone to faults, such as rolling element bearings, gears and rotors. Moreover, a tutorial on how to use SR for machinery fault detection is provided. What’s more, the key issues and prospects of SR in machinery fault detection are pointed out and discussed. It is expected that this review would inspire researchers to explore the potential of SR as well as develop advanced research in this field.