Reliability estimation of complex systems based on a Wiener process with random effects and D-vine copulas. (November 2022)

• Record Type: Journal Article
• Authors: Zheng, Bokai; Chen, Cen; Zhang, Wei; Fu, Rao; Hu, Yifan; Lin, Yigang; Wang, Chunqing; Zhai, Guofu
• Is Part Of: Microelectronics and reliability. Volume 138(2022)
• Publication Date: 2022-11
• Subjects: Wiener process -- Random effects -- D-vine copula -- Reliability estimation -- Dependent performance characteristics (PCs); Electronic apparatus and appliances -- Reliability -- Periodicals; Miniature electronic equipment -- Periodicals; Appareils électroniques -- Fiabilité -- Périodiques; Équipement électronique miniaturisé -- Périodiques; Electronic apparatus and appliances -- Reliability; Miniature electronic equipment; Periodicals; 621.3815
• DOI: 10.1016/j.microrel.2022.114640
• Languages: English
• ISSNs: 0026-2714

Abstract

Abstract: Random effects are ubiquitous in the degradation processes of multiple performance characteristics (PCs) of a complex system, and these PCs are not commonly independent. Neglecting either of the above features will lead to bias in system reliability estimation. Therefore, this study develops a reliability estimation method for the complex systems with multiple PC degradations considering both of the features. First, a Wiener process-based marginal degradation model with random effects (including the random initial PC, unit-to-unit variability, temporal uncertainty, and measurement errors) is proposed to track the degradation path of each PC. A model based on D-vine copulas is put forward to describe the correlation among multiple PCs. Then, a system reliability function is derived based on the two models. Furthermore, a two-step statistical inference is developed to estimate the unknown parameters in the system reliability function. Finally, using a practical case, the effectiveness of the proposed method is illustrated. Highlights: A reliability estimation method is proposed for complex systems. A marginal Wiener process-based degradation model with random effects is proposed. The correlation among multiple performance characteristics is modeled by D-vine copulas. A two-step statistical inference is developed to perform the parameter estimation.