Optimal design for resilient load-sharing systems with nonidentical components

摘要

For system maintenance, strategic component restoration planning is an important conceptual framework for load-sharing G system. A cost-effective treatment of failure events is imperative with the purpose of reinstating the system ability. This paper presents a new optimal design method for load-sharing repairable system, in which a flowgraph is used in conjunction with multiresponse optimization. By introducing the concept of modular design, the system is partitioned into scalable and repairable maintenance modules. The determination of the optimal design depends on the type of system components, the module-based system structure, and the repair rule setting. An extended flowgraph model, which links covariates into transition branches, is used for modeling the system failure evolution. With consideration of various system performance measurements as responses, multiresponse optimization with weighted principal component analysis is used to achieve an optimal design of maintenance modules as well as repair policy. The methodology presented in this paper provides an efficient way to design a system having nonidentical components and arbitrary repair time distributions with consideration of the variety of maintenance policies as well as the diversity of system operating conditions.