Mixed failure-driven and shock-driven mission aborts in heterogeneous systems with arbitrary structure

Failure-driven and condition-driven mission abort and rescue have recently received intensive research attentions in the system reliability field. Little work has addressed both types of aborts simultaneously and the existing model is limited to the special series system structure only. This paper advances the state of the art by modeling mixed failure-driven and condition-driven (specifically, shock-driven) mission aborts in systems with arbitrary system structures and heterogeneous elements. The system performs a primary mission (PM) in random environments exposed to the Poisson process of external shocks. Different system elements have different resistance to shocks, and the shock resistance deteriorates with the number of experienced shocks. A rescue procedure (RP) is triggered to possibly survive the system when a specified number of shocks take place (condition/shockdriven) or when failures of some system elements caused by shocks do not allow continuing the PM execution but allow performing the RP (failure-driven). A universal probabilistic approach is suggested for evaluating the mission success probability (MSP) and system survival probability (SSP) of the considered system. Two example systems (an electric feeder heating system and a smart farm wireless sensor network system) are analyzed. The proposed approach and its applications are demonstrated in determining the optimal mission abort policy that maximizes MSP subject to the SSP constraint or minimizes the expected cost of losses, in the element shock resistance sensitivity analysis, as well as in finding the optimal shock protection replacement solution that minimizes the total cost.

Automated summary

This paper focuses on failure-driven and condition-driven mission abort and rescue in systems with arbitrary structures. The system elements have varying resistance to shocks, and a probabilistic approach is suggested for evaluating mission success probability and system survival probability.