A looseness identification approach for rotating machinery based on post-processing of ensemble empirical mode decomposition and autoregressive modeling

Abstract

The purpose of this research is to investigate the feasibility of utilizing the post-processing of Ensemble Empirical Mode Decomposition (EEMD) and Autoregressive (AR) modeling to identify the looseness faults at different mechanical components of rotating machinery. The post-processing of EEMD is employed to decompose the complicated vibration signals of rotating machinery into a finite number of Intrinsic Mode Functions (IMFs) which represent the mono-oscillated components of different frequency bands. The AR modeling process reveals the fault features of the information-contained IMF components. The characteristic vectors derived from the AR modeling process can be utilized for identifying the locations of looseness faults in the operating machinery. A rotor-bearing test rig is performed to illustrate the looseness faults at different mechanical components. The analysis results show that the proposed approach is capable of identifying the locations of looseness faults at the rotating machinery.

Fault diagnosis of rotating machinery based on time-frequency decomposition and envelope spectrum analysis

“…Additionally, experimental verification can be found in Refs. [16,17]. In this study, each feature is extracted from 4000 points from the vibration signal; hence, the length of the data segment processed by IEEMD is 4000 sampling values.…”

“…All decomposition that results from traditional EMD can be regarded as well-IMFs. Therefore, EEMD and EMD can be combined to resolve the non-IMF problem to develop meaningful components for effective feature extraction based on the methods introduced in the literature [17,18]. According to the decomposition theory of EEMD, the main frequency components in IMFs are lined from high to low, that is to say, the frequency of front IMF is higher than that of latter one.…”

“…According to above analysis, a new IEEMD, consuming less time than these methods in [17] and [18], is proposed for fault diagnosis of the bearings discussed in this paper. This method mainly includes two steps: processing the original vibration signals with traditional EEMD, and using EMD to resolve the decomposition results of EEMD where each step has an IMF selection task.…”

“…Although EEMD has been successfully used to diagnose mechanical faults, its non-IMF problem is a concern and needs to be considered. Therefore, several improved EEMD methods were proposed by Wu [17] and Jiang [18].…”

Multi-Scale Analysis Based Ball Bearing Defect Diagnostics Using Mahalanobis Distance and Support Vector Machine

“…The DI is also known as the Fisher score (FS) and applied to select the signatures for face recognition [33]. The assessment of the Mahalanobis distance (MD) can be used to measure the distinction between the characteristic vectors and demonstrates the effectiveness of identifying the component looseness at different locations of rotating machinery [16,34].…”