Portal de Periódicos da CAPES

CMAFI — Copula-based Multifeature Autocorrelation Fault Identification of rolling bearing

2024; Elsevier BV; Volume: 211; Linguagem: Inglês

10.1016/j.ymssp.2024.111221

1096-1216

Jarosław Duda, Jacek Leśkow, Paweł Pawlik, W. Cioch,

Engineering Diagnostics and Reliability

Diagnosing the rolling bearings using vibration signals requires advanced signal analysis methods. The envelope spectrum analysis method is commonly used, but it requires the determination of the frequency band in which the signal modulation occurs. This paper proposes a method that does not have such limitations and it is based on copula technique applied to wheel bearing signals. Our approach allows to efficiently identify faults using low energy components of the signal estimated via the PCA method. While it is standard to analyze time series autocorrelation function, proposed multifeature autocorrelation analysis finds multiple dominating contribution to joint distribution of values shifted by various time lags. This way it finds also more subtle statistical dependencies, here suggesting damage of bearing. Proposed approach was verified in the laboratory conditions for different kinds of defaults in different positions.