Abstract
Existing reliability models for complex systems mostly focus on system configuration, phased operation, and failure modes, but not on the complicated mission environments. This paper contributes by modeling the system engaged in a primary mission encompassing multiple operation phases with different mission requirements, resource constraints, and random shock environments. The system may fail due to external shocks and the depletion of limited resources, incurring phase-dependent damage costs. To mitigate the risk of system failure, the mission may be aborted before its completion when certain deterioration conditions defined by the number of shocks appears by a certain time. The abort policy may greatly impact the mission success probability (MSP) and expected mission losses (EML). The performance choices of the system in different phases affect resource consumption and task completion time, and consequently also impact MSP and EML. We propose a probabilistic model to derive MSP and EML and further optimize the performance and abort policy to minimize the EML. A case study of a system crossing five zones with four available performance levels is provided to demonstrate the model as well as effects of key parameters (cost, shock rate, shock resistance, performance level, and available resource level) on the mission performance and the optimal aborting and performance policy.