Guidance Navigation and Control Technology Assessment for Future Planetary Science Missions

Beauchamp, Pat; Cutts, J. A.; Quadrelli, Marco B.; Wood, Lincoln J.; Riedel, Joseph E.; McHenry, Mike; Aung, M.; Cangahuala, L. A.; Volpe, Rich

doi:10.2514/6.2013-5411

Cited by 3 publications

(1 citation statement)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Further, while optimization techniques such as the interior-point algorithm can determine an optimal solution for approximately the same computational cost as that associated with building and validating the surrogate model, the surrogate provides more comprehensive information about the system as a whole rather than an isolated optimal solution, including the number and location of local minima, location of the global minimum, and the absolute and relative effects of changes in each of the design parameters on the resulting ∆V . The development of rapid trajectory design and optimization tools such as these were identified by Quadrelli et al (2013), in particular, as a priority for enabling future planetary science missions.…”

Section: Introductionmentioning

confidence: 99%

Reduced cost mission design using surrogate models

Feldhacker

Jones

Doostan

et al. 2016

Advances in Space Research

View full text Add to dashboard Cite

This paper uses surrogate models to reduce the computational cost associated with spacecraft mission design in three-body dynamical systems. Sampling-based least squares regression is used to project the system response onto a set of orthogonal bases, providing a representation of the ∆V required for rendezvous as a reduced-order surrogate model. Models are presented for mid-field rendezvous of spacecraft in orbits in the Earth-Moon circular restricted three-body problem, including a halo orbit about the Earth-Moon L 2 libration point (EML-2) and a distant retrograde orbit (DRO) about the Moon. In each case, the initial position of the spacecraft, the time of flight, and the separation between the chaser and the target vehicles are all considered as design inputs. The results show that sample sizes on the order of 10 2 are sufficient to produce accurate surrogates, with RMS errors reaching 0.2 m/s for the halo orbit and falling below 0.01 m/s for the DRO. A single function call to the resulting surrogate is up to two orders of magnitude faster than computing the same solution using full fidelity propagators. The expansion coefficients solved for in the surrogates are then used to conduct a global sensitivity analysis of the ∆V on each of the input parameters, which identifies the separation between the spacecraft as the primary contributor to the ∆V cost. Finally, the models are demonstrated to be useful for cheap evaluation of the cost function in constrained optimization problems seeking to minimize the ∆V required for rendezvous.These surrogate models show significant advantages for mission design in three-body systems, in terms of both computational cost and capabilities, over traditional Monte Carlo methods.

show abstract

Section: Introductionmentioning

confidence: 99%