Inverse Inference of Component Reliability for k-Out-of-n Systems Based on Maximum Entropy and Hazard-Rate Matrix Representation
Mathematics · 01.04.2026 · 14 (7) · 1181
DOI: 10.3390/math14071181 ↗Kurz gefasst
Fachartikel in der Zeitschrift Mathematics, online erschienen am 01.04.2026. Verfasst von Chao Li und 4 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 49 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Science Challenge Project.
Abstract (Originalsprache)
This study presents a rational inverse inference framework for k-out-of-n systems that derives component-level reliability characteristics from system-level failure and monitoring data. The framework employs a hazard-rate matrix to represent the degradation hierarchy and applies the principle of maximum entropy to allocate system-level probabilities to latent component-state configurations without bias, yielding analytical solutions for component hazard rates. The key innovation lies in combining maximum entropy with the hazard-rate matrix, which overcomes the ill-posed nature of the inverse problem and enables systematic integration of heterogeneous auxiliary information within a unified formulation, including system-level multi-state observations, component-wise moment constraints, sub-component data, and inter-component dependencies. This flexibility addresses a major limitation of existing inverse methods, such as Bayesian approaches, which are typically restricted to a single data type and often require strong prior assumptions or extensive failure datasets. The practical applicability of the framework is demonstrated through a case study of a west-to-east gas pipeline pumping system, highlighting its effectiveness in processing multiple information types and delivering actionable component-level reliability assessments for maintenance decision support. To the best of our knowledge, this is the first study to formulate and solve the inverse inference problem for k-out-of-n systems in a theoretically grounded and information-theoretically optimal manner.
Förderung
- Science Challenge Project
Zitieren
APA
Li, C., Gong, T., Zhou, D., He, J., & Guan, X. (2026). Inverse Inference of Component Reliability for k-Out-of-n Systems Based on Maximum Entropy and Hazard-Rate Matrix Representation. Mathematics, 14(7), 1181. https://doi.org/10.3390/math14071181
Vancouver
Li C, Gong T, Zhou D, He J, Guan X. Inverse Inference of Component Reliability for k-Out-of-n Systems Based on Maximum Entropy and Hazard-Rate Matrix Representation. Mathematics. 2026;14(7):1181. doi:10.3390/math14071181
BibTeX
@article{Li2026,
author = {Li, Chao and Gong, Tianci and Zhou, Daoqing and He, Jingjing and Guan, Xuefei},
title = {Inverse Inference of Component Reliability for k-Out-of-n Systems Based on Maximum Entropy and Hazard-Rate Matrix Representation},
journal = {Mathematics},
year = {2026},
volume = {14},
number = {7},
pages = {1181},
doi = {10.3390/math14071181}
}
EUVIMED