Files in this item

FilesDescriptionFormat

application/pdf

application/pdfELLISON-DISSERTATION-2018.pdf (18MB)
(no description provided)PDF

Description

Title:Robust preliminary design for multiple gravity assist spacecraft trajectories
Author(s):Ellison, Donald Hamilton
Director of Research:Conway, Bruce A.
Doctoral Committee Chair(s):Conway, Bruce A.
Doctoral Committee Member(s):Prussing, John E.; Hirani, Anil N.; Coverstone, Victoria L.; Englander, Jacob A
Department / Program:Aerospace Engineering
Discipline:Aerospace Engineering
Degree Granting Institution:University of Illinois at Urbana-Champaign
Degree:Ph.D.
Genre:Dissertation
Subject(s):Sims-Flanagan
Low-Thrust
Trajectory
Optimization
Spacecraft
Orbit
Sparse Nonlinear OPTimizer (SNOPT)
Automatic Differentiation
Jacobian
Partial Derivatives
State Transition Matrix
Nonlinear Program
Moon Tour
Monotonic Basin Hopping
Evolutionary Mission Trajectory Generator (EMTG)
Sundman
Finite-Burn
MIRAGE
OpenMP
Parallel Computing
Abstract:The development of a new spacecraft trajectory design method most often occurs because a particular capability does not exist. The invention is usually considered successful so long as it is capable of producing solutions to the problem in question, and thus satisfies a particular design requirement, or is mathematically elegant. When innovation favors the latter to the exclusion of the former, the interoperability of the new method with existent techniques, and the utility of the method in the context of the overall mission design process, from concept to fight operations, is not always realized. The concepts introduced and developed throughout this work respond to specific preliminary mission design needs, but their development is also focused on maintaining, or improving, trajectory design work flow compatibility and efficiency. The techniques described address specific contributions to the multiple gravity assist trajectory optimization state-of-the-art, however, each one is also an important component of a modern trajectory design paradigm and is valuable for its ability to be integrated with and streamline that process as a whole. The bounded-impulse approximation is a widely used method for early stage trajectory design for low and high thrust vehicles. Many previous studies involving this method of design have focused on developing new or improved trajectory transcriptions. This dissertation introduces analytic techniques for calculating the Jacobian matrix for two existing bounded-impulse trajectory models. The calculations allow for the use of a smooth spacecraft power model. One such model is introduced that handles multiple thruster on-off events and a variety of logic programs. A smooth spline-based ephemeris system is also discussed that is compatible with the analytic Jacobian formulae. Mission design activities associated with the NASA New Frontiers 4 proposal call identified a particular shortcoming of the popular Sims-Flanagan bounded-impulse model, namely that its control nodes are distributed equally in time, clustering them at apoapsis for eccentric transfers, which reduces the control authority at periapsis. In response to this, a partially regularized bounded-impulse model is introduced that distributes control nodes equally at both apses. The new transcription is capable of delivering the same mass to the target as a trajectory modeled using Sims-Flanagan, but requires far fewer control segments in the trajectory discretization to do so. A bounded-impulse trajectory is usually sufficient to generate a first order (or better) estimate of a mission's mass budget, however, these low-fidelity models can prove problematic to converge inside a flight-fidelity design tool that models fi nite-burn arcs. Low-thrust trajectories in particular suffer from this issue due to the extended period of time that the thruster operates. Analytic partial derivative computations are introduced in this work that enable the replacement of bounded-impulse maneuvers and Keplerian propagation with numerical integration arcs without reducing the runtime performance such that the model becomes unusable for preliminary design tasks. These calculations are also compatible with a smooth electric power model and accommodate final time free problems. The resulting trajectory is shown to be sufficiently accurate such that the flight heritage tool MIRAGE can track it within acceptable error limits placed on the spacecraft's state vector. Finally, improvements that this thesis makes to bounded-impulse trajectory models are leveraged to solve a challenging planetary satellite tour design problem. The robustness improvements that the analytic Jacobian formulae afford the trajectory transcription, are combined with a parallelized flyby tree path finding algorithm to produce a design framework capable of autonomously optimizing a moon tour mission similar to Galileo from launch through tour end using two-body dynamics.
Issue Date:2018-07-10
Type:Text
URI:http://hdl.handle.net/2142/101551
Rights Information:Copyright 2018 by Donald Hamilton Ellison. All rights reserved.
Date Available in IDEALS:2018-09-27
Date Deposited:2018-08


This item appears in the following Collection(s)

Item Statistics