Predicting remaining useful life of an individual unit using proportional hazards model and logistic regression model

• Published in: RAMS '06. Annual Reliability and Maintainability Symposium, 2006 pp. 127 - 132
• Main Authors: Haitao Liao, Wenbiao Zhao, Huairui Guo
• Format: Conference Proceeding
• Language: English
• Published: IEEE 2006
• Subjects: Civil engineering; Degradation; Engines; Feature extraction; Hazards; Logistics; Predictive models; Sensor phenomena and characterization; Signal processing; Turbines
• ISBN: 1424400074; 9781424400072
• ISSN: 0149-144X; 2577-0993
• DOI: 10.1109/RAMS.2006.1677362

Reliability of an individual unit during field use is important in many critical applications such as turbine engines, life-maintaining systems and civil engineering structures. The remaining useful life (RUL) of the unit indicates its ability of surviving the operation in the future. When the failure indication (degradation) has been detected, it is essential to estimate the RUL accurately for making a timely maintenance decision for failure avoidance. In recent years, RUL prediction in service has received increasing attention. As many powerful sensors and signal processing techniques appear, multiple degradation features can be extracted for degradation detection and quantification. These features can serve as the basis for RUL prediction. This paper presents the proportional hazards model and logistic regression model, which relates the multiple degradation features of sensor signals to the specific reliability indices of the unit, and enable us to predict its RUL. Comparisons are made for the two models regarding their effectiveness and computation effort. An example of bearing test is provided to demonstrate the proposed approach in practical use. The results show that the models are capable of providing accurate RUL prediction to support timely maintenance decisions