Multi-objective optimization of the fleet mix problem using the SaFER model

One of the most important tasks for an organization which transports cargo and people is the determination of number and type of platforms which will be needed. Due to the presence of multiple conflicting objectives, such as cost and performance, this problem may be considered multi-objective. In order to estimate the fleet that can fulfill the scenario requirements, the Stochastic Fleet Estimation - Robust (SaFER) model was previously developed. It uses scheduling heuristics and optimization. However, using the SaFER model within a multi-objective optimization framework is not computationally feasible; therefore, a surrogate model is proposed in this paper to approximate SaFER for use in the fitness evaluations of schedule cost objectives. An artificial military air mobility dataset is used to demonstrate the increase in speed of the surrogate model over SaFER, and the accuracy of the surrogate model in estimating schedule costs versus SaFER.