Optimal Replacement Policy for Multi-State Resilient Systems under Non-Fatal Random Shocks  
Author

Yong-Hao Lin

 

Co-Author(s)

Ruey-Huei Yeh

 

Abstract Most reliability models for multi-state systems (MSS) assume that external shocks result in immediate system failure. However, modern resilient systems frequently experience soft failures—non-fatal shocks that degrade system performance without causing total shutdown. In such cases, the system continues operating at a reduced performance level while undergoing self-recovery. This study develops a comprehensive
reliability framework incorporating internal degradation, external random shocks, and resilience mechanisms. A continuous-time Markov chain (CTMC) is constructed to characterize system state transitions, introducing the concept of “Virtual Operable States” to model soft failure dynamics explicitly. A cost model is further
developed to evaluate the long-run expected cost rate and determine the optimal preventive replacement threshold. The proposed framework provides a realistic analytical foundation for evaluating systems with complex self-recovery capabilities. Numerical results demonstrate that neglecting non-fatal shocks leads to overestimating availability and underestimating lifecycle
costs.

 

Keywords Replacement policy, Continuous-time Markov chain, Non-fatal random shock, Resilience
   
    Article #:  RQD2026-137
 

Proceedings of 31st ISSAT International Conference on Reliability & Quality in Design
August 5-7, 2026