Multi-Objective Optimization Of Fleet-Level Metrics To Determine New System Design Requirements: An Application To Military Air Cargo Fuel Efficiency

Abstract : In many instances, yet-to-be-acquired military systems will, when deployed, operate alongside existing systems to provide a set of capabilities. The design requirements assigned to the yet-to-be-designed systems will impact the performance with which the resulting system of systems provides the desired capabilities. Identifying these design requirements for new, yet-to-be introduced systems is difficult, because quantifying the impact of these requirements on fleet level metrics must use some sort of analyses that recognizes the tight coupling of the system design problem (in this paper, a new aircraft design) and asset assignment problem (in this paper, aircraft fleet deployment to provide military cargo transportation). The methodology presented here addresses this by solving a combined platform design, fleet operations and acquisition-level decision-making problem wherein the design requirements of a new system (or systems) appear as design variables in an optimization problem formulation. The approach employs a decomposition strategy to describe the design requirements of the new, yet-to-be-acquired system so that the new system improves fleet-level performance. This fleet-level performance usually involves multiple, competing fleet-level objectives; the research investigates tradeoffs between objectives of fuel usage (cost) and fleet-wide productivity using a relatively simple example motivated by the USAF Air Mobility Command cargo carrying aircraft fleet. Solutions to the multi-objective optimization problem represent the best possible tradeoffs of these two objectives as functions of the new aircraft design requirements. Presenting these results as a Pareto frontier shows the relationship of fleet productivity to fuel cost; this has features of a fuel cost as an independent variable context for decision-making.