Previous work in air transportation system-of-systems (ATSoSs) design optimization considered integrated aircraft sizing, fleet allocation, and route network configuration. The associated nested multidisciplinary formulation posed a numerically challenging blackbox optimization problem; therefore, direct search methods with convergence properties were used to solve it. However, the complexity of the blackbox impedes greatly the solution of larger-scale problems, where the number of considered nodes in the route network is high. The research presented here adopts a rule-based route network design inspired by biological transfer principles. This bio-inspired approach decouples the network configuration problem from the optimization loop, leading to significant numerical simplifications. The usefulness of the bio-inspired approach is demonstrated by comparing its results to those obtained using the nested formulation for a 15 city network. We then consider introduction of new aircraft as well as a larger problem with 20 cities.
Uncertainty-Aware Search Framework for Multi-Objective Bayesian Optimization
