An Optimization-Based Approach to Determine System Requirements Under Multiple Domain-Specific Uncertainties

Abstract

The task of determining the optimal design requirements of a new system, which will operate along with other existing systems to provide a set of overarching capabilities, is challenging due to the tightly coupled effects that setting requirements on a system's design can have on how the operator uses the system. In this paper, the new system is a strategic military cargo aircraft and the other systems are a fleet of different, existing cargo aircraft; a subset of actual fleet operations from the U.S. Air Force Air Mobility Command defines the example problems in this work. This research builds upon prior efforts to develop a quantitative approach that identifies optimum design requirements of new, yet-to-be-designed systems that, when serving alongside other systems, will optimize fleet-level objectives. The new efforts here address the effect of various uncertainties. The approach incorporates techniques from multidisciplinary design optimization, statistical theory, and robust/reliability-based methods to develop computationally tractable approaches for this kind of problem. The paper also demonstrates the ability to generate tradeoffs between a cost-related metric of fleet-level fuel usage and a performance related metric of fleet-wide productivity. A possible extension for application in commercial air travel also appears in the paper.