Multi-stage stochastic engine usage optimization for fighter jet fleet using nested decomposition algorithm

Fighter jet squadrons face a critical challenge: meeting rigorous monthly flight-hour targets while managing strict engine service life limitations. This complex task necessitates the optimal allocation of engine resources and meticulous planning of flight hours for each aircraft, thereby balancing operational demands with maintenance imperatives. Our study addressed this multifaceted challenge by proposing a novel multi-stage stochastic programming (MSSP) model. Under uncertainty considerations, the model assists engine maintenance contractors in determining when to disassemble and reassemble fighter jet engines to ensure fighter jets meet the flight-hour requirements of the air force. Unlike previous deterministic approaches, our model incorporates random factors and uncertainties inherent in aviation operations, such as weather variability and mission changes. This comprehensive approach represents a considerable advancement in the field. To tackle the exponential increase in problem complexity at practical scales, we developed a nested decomposition algorithm. This innovative algorithm efficiently decomposes large-scale problems into manageable subproblems, utilizing tight lower bounds and problem-specific cuts to enhance computational efficiency. Empirical studies based on real world planning settings show that, when compared with existing manual planning practices, the proposed approach reduces the number of engines reaching their service life limits by 15.3 percent and increases available flight hours by 465.66 hours, thereby demonstrating clear and substantial operational benefits.