Jixie chuandong (Jan 2016)
Rolling Bearing Performance Degradation Assessment based on the Wavelet Packet Tsallis Entropy and FCM
Abstract
The rolling bearing performance degradation assessment method is proposed by the joint use of wavelet packet Tsallis entropy（ WPTE） and fuzzy C- means（ FCM）. Firstly,the WPTEs of normal data and final failure data are extracted followed by the evaluation model of performance degradation being built by using FCM. Then,the degree of tested data belonging to the final failure state is employed as an index to evaluate the level of bearing performance degradation in a quantitative way. Meanwhile,a threshold is set to alert incipient faults. Experimental analysis shows that the proposed index of bearing performance degradation is able to detect bearing fault in its early stage and has stronger robustness to the noise in comparison with the assessment method based on wavelet packet entropy and FCM. In addition,the proposed index has a value located in [0,1]and a reasonable interpretation to the evaluated bearing health condition.
