Flexible reliability assessment model for self-recovery systems with performance degradation

LI Wenxue; TANG Jiayin; SONG Qingbiao

doi:10.16579/j.issn.1001.9669.2026.03.003

Abstract：[Objective]Aiming at the problem that complex systems such as gear transmission have the ability to recover from performance degradation damage after external random shocks,and existing models are difficult to accurately evaluate reliability,a flexible reliability assessment model was constructed from the perspective of competitive failure to realize precise reliability analysis of such systems.[Methods]Firstly,a nonlinear variable-rate Wiener process was adopted to characterize performance degradation,considering random factors and individual differences to provide a basis for describing dynamic degradation characteristics.Secondly,a Poisson process was coordinated to characterize random shocks,clarifying the influence mechanism of fatal and non-fatal shocks.Then,a self-recovery function was used to describe the complete or partial recovery process of degradation damage,and a flexible competitive failure reliability model was constructed.Finally,a multi-stage parameter estimation method was proposed based on the quasi-Newton iterative algorithm and cubic spline interpolation to ensure the accuracy of model parameter solution.[Results]Example verification shows that after the system operates for about 1 100 hours,ignoring the self-recovery effect will underestimate the reliability level.The reliability difference between the established model and the simulation results does not exceed 0.015 7,and the model is applicable to linear/constant degradation rate systems and classic systems without self-recovery,providing a reference for the reliability assessment of self-recovery systems such as gear transmission.