A Fast, Robust Genetic Algorithm for Producing Families of Constrained Multi-Burn Orbit Transfers

Eric George, The Aerospace Corporation

Keywords: trajectory,optimization,maneuver

Abstract:

“Pork Chop” contour plots of Delta-V as a function of departure and arrival times, initially used for planning interplanetary trajectories, are becoming an important tool for the visualization and analysis of geocentric satellite maneuver envelopes.
The prototypical pork chop plot is based on single burn Lambert transfers for each departure-arrival time pair. However, these transfers are often very inefficient for scenarios involving a significant plane change and have little ability to satisfy additional constraints.
This paper presents a novel Genetic Algorithm approach for producing optimal multi-burn, non-coplanar trajectories for all departure-arrival time pairs in a single run using a distributed population with localized selection and replacement operations. The algorithm determines, for each departure-arrival time pair, both number of maneuvers and the timing, direction and magnitude for each maneuver, with the foundational objective of minimizing the overall Delta-V.
This technique has also proven adept at minimizing Delta-V, subject to additional constraints such as lighting conditions and relative velocity. The sample scenario presented improves upon the Lambert solution by a factor of up to 8 and illustrates the ability of the approach to open large areas of the solution space previously excluded by a lighting constraint.

Date of Conference: September 19-22, 2023

Track: Astrodynamics

View Paper