A Vibration Acceleration Sensor-based Shock Pulse Method for Condition Evaluation of Rolling Bearings

Compared to the acceleration sensor based vibration measurement, the shock pulse sensor based SPM has been proved to be simpler and more effective for evaluating rolling bearings’ condition, but it also has the inevitable inferiority of being unable to monitor the overall state of the machine. To solve this problem, a novel shock pulse method based on acceleration sensor (APM) is proposed in this paper. It widens the function of an acceleration sensor and realizes the analysis of resonance response beyond the linear response of the sensor. This paper focuses on two core issues. The first one is to extract the available shock pulse from vibration acceleration signals by the band-pass filtering and envelope demodulation. The key to this issue is the bandwidth selection of the band-pass filter. In this study, the deviation rate of pulse value is defined, and the bandwidth is optimized by dealing with the contradiction between the deviation rate and the anti-noise ability of the filter. Another issue is the measurement of the pulse strength. In this paper, an improved pulse quantization method is proposed to reduce the interference of the random impact. Furthermore, the normalization and condition estimation of SPM theory are used to realize the mapping between the measurement value and the bearing’s condition. Experimental bearing signals with various fault status were analyzed to verify the feasibility and effectiveness of APM.