Bearing Performance Degradation Assessment by Orthogonal Local Preserving Projection and Continuous Hidden Markov Model

Abstract

Bearing is the key component in rotating machine. It is important to assess the performance degradation degree of bearings for making proactive maintenance and realizing near-zero downtime. A methodology based on orthogonal local preserving projection (OLPP) and continuous hidden Markov model (CHMM) is introduced in bearing performance degradation assessment. Firstly, the time domain, frequency domain and time-frequency domain features are extracted from the vibration signals. Then, the multi-dimensional features are reduced by OLPP. And the selection of the adjacent paragraph parameters in OLPP is optimized adaptively by minimizing the ratio of between-class distance to within-class distance. A CHMM is trained by using the reduced feature in normal condition. At last, the test bearing data are input into the pre-trained CHMM to calculate the log-likelihood of the test data, which can assess the performance degradation of bearings quantitatively. A bearing accelerated life experiment is performed to validate the feasibility and validity of the proposed method.

Challenges and Opportunities of AI-Enabled Monitoring, Diagnosis & Prognosis: A Review

“…To establish the index with an obvious trend, Li et al [225] proposed the negative log likelihood probability based on the two-dimensional HMM as the bearing performance degradation index, showing the sensitivity to weak defects. Liu et al [226] proposed a bearing DA method based on orthogonal local preserving projection (OLPP) and continuous HMM. The continuous HMM was used to train the data after dimension reduction by OLPP, and then the performance could be evaluated quantitatively by calculating the logarithmic likelihood of the data.…”

A Fuzzy Performance Evaluation Model for a Gearbox System Using Hidden Markov Model

“…Wang et al [24] identified the running state of the spot welding machining process by HMM. Liu et al [25] proposed a bearing performance degradation evaluation method by orthogonal locally retained projection and continuous HMM. Zhou et al [26] proposed a fault diagnosis by a shift-invariant dictionary learning and HMM.…”