Multi-attempt missions with multiple rescue options

In spite of intensive works on modeling and optimizing diverse mission aborting systems from the reliability community, none of the existing models have considered multiple rescue options. This paper pioneers the study of a multi-attempt mission system with different types of rescue procedures characterized by dissimilar system performance, shock rates, and costs. A new optimization problem is formulated, which determines the aborting rule and rescue option choice for each attempt to minimize the expected mission losses (EML) encompassing the mission failure penalty, system loss cost and rescue cost. A new numerical algorithm is then put forward to assess the EML of the considered multi-attempt mission system under any given mission abort and rescue policy (MARP). The genetic algorithm is further implemented to solve the MARP optimization problem. The proposed model is demonstrated through an unmanned aerial vehicle (UAV) performing a payload delivery mission with four possible rescue options. The UAV case study also investigates the accuracy and efficiency of the proposed numerical EML evaluation algorithm and reveals the impacts of several key parameters (mission failure penalty, system loss cost, the allowed number of attempts, shock resistance deterioration factor) on the EML and the optimal MARP solutions.