A bearing reliability assessment method based on an improved Hidden Markov Model

In the past decade, significant progress has been made in the reliability research of CNC machine tools to enhance their credibility among users. Mechanical components in such equipment often exhibit performance degradation characteristics, leading to a decline in reliability over time. Typically, reliability data of equipment is obtained through statistical analysis of a large amount of test and operation data. However, CNC machine tools, being long-lived and expensive large-scale equipment, generally face the problem of insufficient on-site operation reliability data samples, which makes reliability assessment difficult and prone to frequent machine failures. To address the issue of small sample data, this paper proposes a reliability assessment method based on the Hidden Markov Model (HMM) and considering performance degradation - the Multivariate Dynamic Degradation Hidden Markov Model (MDD-HMM). This method first uses the monitoring data of performance parameters to deduce the performance degradation patterns of key components of the machine tool, such as bearings. Then, these performance degradation patterns are incorporated into the Hidden Markov Model to construct a time-related state transition probability matrix that can represent the performance degradation process. Based on the state probabilities calculated from the degradation Hidden Markov Model, the reliability curve of the equipment can be derived. Finally, the reliability assessment was conducted using the full life cycle data of IMS bearings, and the results verified the effectiveness of the model under small sample conditions.