In this paper we present a new robust approach to produce efficient aircraft using numerical optimization. Our focus is on performing a multi-point optimization that considers the performance at multiple operating points simultaneously. The goal is to avoid severe performance degradation at off-design conditions, which typically occurs with a single-point optimization. Specifically, we aim to design a fuel-efficient long-range aircraft configuration. The robustness is introduced by considering hundreds of missions within the operational flight envelope of similar sized aircraft, based on historical data for the actual flight operations. Due to the large computational cost associated with the high-fidelity multidisciplinary analysis, kriging surrogate models are employed to allow thousands of detailed flight analyses to be performed while limiting the number of high-fidelity evaluations. The methodology is demonstrated in a fuel burn minimization problem of a long-range wide-body aircraft configuration.
展开▼
