Multivariate Storage Degradation Modeling Based on Copula Function

Li, Yong; Xue, Peng

doi:10.1155/2014/503407

Cited by 10 publications

(7 citation statements)

References 13 publications

(13 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We set six different combinations of (m, N): (10, 8), (10,15), (10,25), (20,8), (20,15), (20,25), where m represents the number of observations, and N represents the number of samples. Then, each of the combinations is simulated 400 times with the same parameter settings as the simulation study, and the simulated degradation data are used for parameter estimation.…”

Section: Fundingmentioning

confidence: 99%

See 1 more Smart Citation

Degradation modeling and reliability prediction of products with indicators influenced by clusters in a dynamic environment

Huang

Zhou³

et al. 2022

Proceedings of the Institution of Mechanical Engineers, Part O:

View full text Add to dashboard Cite

Modern products have tended to gain increasingly complex structures, and most of them have multiple dependent performance indicators (PIs), which makes the degradation modeling and reliability prediction of such products challenging. For some products, based on the design of the products, their PIs are influenced by some common underlying components called clusters in this paper, thus they are correlated. Besides, in engineering practice, products may suffer from a dynamic environment that causes difficulties in reliability analyses. To address these situations, this paper proposes a multivariate degradation model based on the Wiener process and establishes the correlation among the PIs in a flexible and intelligible way. The effect of the dynamic environment on the degradation rate is considered to be a multiplication form. The expectation-maximization (EM) algorithm is utilized to achieve an accurate estimation of the model parameters. The expression of the reliability function of the product is obtained by a tangent approximation approach. In the end, a simulation study and a case study are presented to demonstrate the effectiveness and application of the proposed model.

show abstract

Section: Fundingmentioning

confidence: 99%

“…19 When multiple PIs of a product degrade at the same time, it is not appropriate to predict the lifetime of the product based on a model that only considers one of the PIs, as it would lead to inaccurate results. 20…”

Section: Introductionmentioning

confidence: 99%

Degradation modeling and reliability prediction of products with indicators influenced by clusters in a dynamic environment

Huang

Zhou³

et al. 2022

Proceedings of the Institution of Mechanical Engineers, Part O:

View full text Add to dashboard Cite

show abstract

“…In the literature, Ye et al [167] compared the goodness-of-fitting of four different types of Wiener process-based models, namely, (1) the simple Wiener process; (2) the RD-Wiener model, in which the drift coefficient is randomly-distributed; (3) the RV-Wiener model where the diffusion coefficient is treated as a random variable; and (4) the RDV-Wiener model, in which both the drift and diffusion coefficient are normally-distributed variables. More recently, the normal distribution is replaced with the skewnormal distribution since the former one can bring biases in parameter estimation [171], [174], [175]. Though the skewnormal-distributed drift coefficient can improve the accuracy of assessment results, it can also lead to the difficulty in other aspects of the statistical analysis of degradation data [174].…”

Section: A: Wiener Processmentioning

confidence: 99%

Modeling and Analysis of Performance Degradation Data for Reliability Assessment: A Review

Chen

Liu

et al. 2020

IEEE Access

View full text Add to dashboard Cite

Reliability engineering plays an important role in the design, manufacture, maintenance, and replacement of industrial products. Over the last few decades, accelerated degradation testing (ADT) has been largely utilized to shorten test durations, reduce the samples needed, and provide sufficient degradation data to ensure the effective reliability assessment of the concerned products. Meanwhile, performance degradation modeling has been recognized as an essential approach to help researchers and producers understand the health conditions of the deteriorating systems. However, the diversity in reliability tests, degradation models, and statistical analysis techniques has increased the difficulty in selecting appropriate reliability assessment methods in specific scenarios. Besides, there are no systematic reviews focused on modeling and analysis of performance degradation data. Therefore, this paper aims to (1) present ADT fundamentals, including the basic theory, ADT methods, accelerated stress variables, type of acceleration models, as well as ADT optimization, (2) comprehensively review current states and future challenges in degradation modeling, (3) discuss the problem of model mis-specification and compare different approaches for parameter estimation, (4) highlight future opportunities and possible directions deserving further research.

show abstract

“…Although some studies have investigated the dependency between the failure modes, a key problem with much of the literature regarding the dependency between the modes is that they just considered dependency when the modes are of the same type and the distributions of the modes are the same (Li and Hao, 2016;Xiaogang and Peng, 2014). The current study developed a methodology that took into consideration the dependency between both degradation measurements (soft failure) and observed failure time (hard failure) when the distribution of the failure modes is different.…”

Section: Introductionmentioning

confidence: 99%

Copula-based reliability analysis of gamma degradation process and Weibull failure time

Mireh

Khodadadi

Haghighi

2019

IJQRM

View full text Add to dashboard Cite

Purpose The purpose of this paper is the reliability analysis for systems with dependent gamma degradation process and Weibull failure time. Design/methodology/approach Consider a life testing experiment in which a sample of n devices starts to operate at t=0 and the data are available on failure time and failure-evolving process on each individual, called in some contents wear or degradation. Ignoring the between performance characteristics dependency structure may lead us to different reliability estimations, while the dependency justly exists. In previous research, dependency between the degradation process and hard failure time has been studied in limited detail (special closed form expression). Thereafter, the dependency between two degradation processes with the same structure (gamma process) in a system is considered using the copula function. Findings The results indicate that ignoring the dependency structure may lead us to different reliability estimations while the dependency justly exists. Originality/value This study gives some contributions that evaluate reliability metrics with more than one failure mechanism that may not be independent and possibly follow a different distribution function. The authors have used the copula function as a basis to develop a proposal model and analysis methods. In addition, the authors discussed the identifiability of the copula. Finally, simulation data were used to review the suggested approach.

show abstract

Multivariate Storage Degradation Modeling Based on Copula Function

Cited by 10 publications

References 13 publications

Degradation modeling and reliability prediction of products with indicators influenced by clusters in a dynamic environment

Degradation modeling and reliability prediction of products with indicators influenced by clusters in a dynamic environment

Modeling and Analysis of Performance Degradation Data for Reliability Assessment: A Review

Copula-based reliability analysis of gamma degradation process and Weibull failure time

Contact Info

Product

Resources

About