Search for Author, Title, Keyword
RESEARCH PAPER
Remaining useful life prediction of binary stochastic degradation equipment based on mixed Copula functions
,
 
,
 
,
 
,
 
 
 
More details
Hide details
1
School of Mechanical and Electrical Engineering, Lanzhou University of Technology, China
 
2
School of Mechanical and Electrical Engineering, Lanzhou University of Technology, Lanzhou 730050, China, China
 
 
Submission date: 2025-03-21
 
 
Final revision date: 2025-06-13
 
 
Acceptance date: 2025-08-25
 
 
Online publication date: 2025-09-07
 
 
Publication date: 2025-09-07
 
 
Corresponding author
Jianhua Li   

School of Mechanical and Electrical Engineering, Lanzhou University of Technology, Lanzhou 730050, China, China
 
 
 
HIGHLIGHTS
  • A remaining useful life prediction method based on mixed Copula functions is proposed.
  • A stochastic degradation model considering multiple degradation modes is established.
  • Comparative analysis shows that the proposed method results in smaller prediction errors.
KEYWORDS
TOPICS
ABSTRACT
In view of the problem of distorted characterization of correlated performance features in the existing RUL prediction methods for degrading equipment based on multivariate correlation characteristics, which is caused by relying on a single Copula function to model the relationships among multiple degradation features, this paper proposes an RUL prediction method based on a mixed Copula function.First, a stochastic degradation model considering multiple degradation modes is established based on the Wiener process. Second, a mixed Copula function formed by a linear combination of Gumbel, Clayton, and Frank Copulas is constructed to characterize the complex correlations among degradation features. Furthermore, the step - by - step maximum likelihood method is used to estimate the parameters in the model. Finally, the effectiveness of the proposed method is verified through simulated degradation datasets, degradation data of LED lighting systems, and metal crack degradation data.
eISSN:2956-3860
ISSN:1507-2711
Journals System - logo
Scroll to top