Simultaneous Design and Trajectory Optimization for Boosted Hypersonic Glide Vehicles
Jonathan R. Cangelosi, Matthias Heinkenschloss, Jacob T. Needels, Juan J. Alonso
Abstract
This manuscript describes a methodology for simultaneous vehicle and trajectory optimiza- tion of a hypersonic glide vehicle. The co-design problem is formulated as an optimization problem with constraints including vehicle dynamics, path constraints (e.g., surface heating), and other constraints. The discretized optimization problem is solved simultaneously in the vehicle design parameters, the state variables, and the controls using an interior point method. Gaussian process (GP) surrogates, which are generated from sample candidate designs and flight conditions, are used to model vehicle aerodynamic performance and mass properties, as well as their first and second-order derivatives required by the optimizer. These GP surrogates and their derivatives are computationally inexpensive, making the all-at-once optimization approach for the co-design problem more tractable. To mitigate the effect of surrogate model errors on the solution of the optimal control problem, the GP models are refined using samples of the vehicle aerodynamic performance and mass properties at the solution of the co-design problem with the current surrogate. The resulting framework is applied to maximizing the range of a hypersonic glide vehicle with path and terminal constraints. Possible extensions of this methodology are also discussed, including the incorporation of more complex vehicle models such as multi-fidelity models, as well as adaptive surrogate modeling strategies to mitigate the effect of model errors on the solution of the optimal control problem.