Reliability allocation method based on minimizing implementation risk

• Published in: Expert systems with applications Vol. 255; p. 124746
• Main Authors: Axita, Yang, Zhaojun, Chen, Chuanhai, Guo, Jinyan, Yang, Shang
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.12.2024
• Subjects: Belief reliability theory; Optimization model; Reliability allocation; Risk function; Risk influencing factors; Risk function; Belief reliability theory; Reliability allocation; Optimization model; Risk influencing factors
• ISSN: 0957-4174
• DOI: 10.1016/j.eswa.2024.124746

Reliability allocation is an important part of product design. Whether the reliability allocation result is reasonable or not is directly related to the smooth development of the implementation process. This study proposes a new reliability allocation method to overcome the limitations of traditional reliability allocation methods. These limitations include the idea that system reliability cannot be reasonably described when there is epistemic uncertainty, and the potential implementation risks of reliability allocation results are not considered. Firstly, the reliability of each unit is described by a random basic unit and an uncertain basic unit, thereby improving the disadvantage that the existing methods cannot reasonably describe the reliability of the unit with epistemic uncertainty. Moreover, system reliability is measured based on belief reliability theory, which can obtain a reasonable system reliability. Secondly, an implementation risk function is constructed by considering technology readiness, comprehensive strength of the supplier and failure correlation as risk influencing factors. On this basis, a reliability optimization allocation model is established with the implementation risk function as optimization objective. Moreover, the reliability allocation result with minimum implementation risk is obtained by solving the optimization model. Finally, the advantages of the proposed method are illustrated by a case analysis.