Application of the largest Lyapunov exponent algorithm for feature extraction in low speed slew bearing condition monitoring. (January 2015)

• Record Type: Journal Article
• Authors: Caesarendra, Wahyu; Kosasih, Buyung; Tieu, Anh Kiet; Moodie, Craig A.S.
• Is Part Of: Mechanical systems and signal processing. Volume 50/51(2015)
• Page Start: 116
• Page End: 138
• Publication Date: 2015-01
• Subjects: Empirical mode decomposition -- Feature extraction -- Largest Lyapunov exponent -- Low speed slew bearing; Structural dynamics -- Periodicals; Vibration -- Periodicals; Constructions -- Dynamique -- Périodiques; Vibration -- Périodiques; Structural dynamics; Vibration; Periodicals; 621
• DOI: 10.1016/j.ymssp.2014.05.021
• Languages: English
• ISSNs: 0888-3270

Abstract

Abstract: This paper presents a new application of the largest Lyapunov exponent (LLE) algorithm for feature extraction method in low speed slew bearing condition monitoring. The LLE algorithm is employed to measure the degree of non-linearity of the vibration signal which is not easily monitored by existing methods. The method is able to detect changes in the condition of the bearing and demonstrates better tracking of the progressive deterioration of the bearing during the 139 measurement days than comparable methods such as the time domain feature methods based on root mean square (RMS), skewness and kurtosis extraction from the raw vibration signal and also better than extracting similar features from selected intrinsic mode functions (IMFs) of the empirical mode decomposition (EMD) result. The application of the method is demonstrated with laboratory slew bearing vibration data and industrial bearing data from a coal bridge reclaimer used in a local steel mill. Highlights: This paper presents a new application of the largest Lyapunov exponent (LLE) algorithm for feature extraction. The increase deterioration level of slew bearing is detected. The comparison between LLE feature and time-domain features and EMD is reported. The most important LLE parameters namely reconstruction delay J is selected based on the FFT.