Gregory Badura, Georgia Tech Research Institute; Yuri Shimane, Georgia Institute of Technology; Lois Visonneau, Georgia Tech; Matthew Gilmartin, Georgia Institute of Technology; Stef Crum, Georgia Institute of Technology; Christopher Valenta, Georgia Tech Research Institute; Micheal Steffens, Georgia Institute of Technology; Selcuk Cimtalay, Georgia Institute of Technology; Francis Humphrey, Georgia Institute of Technology; Mariel Borowitz, Office of Space Commerce; Brian Gunter, Georgia Institute of Technology; John Christian, Georgia Institute of Technology; Koki Ho, Georgia Institute of Technology
Keywords: Cislunar SSA, MDAO
Abstract:
Performing Space Domain Awareness (SDA) on the cislunar volume is infeasible with any single Electro-Optical or Infrared (EO/IR) sensor system. This difficulty results from factors such as the vast expanse of cislunar space, the chaotic dynamic conditions, and the time-varying illumination backgrounds across the volume. These complications necessitate the development of collaborative networks of distributed space-based sensors to close coverage gaps of existing Earth-based infrastructure and enable persistent cislunar SDA.
We show progress on optimizing the distributed sensor optimization problem from the lens of not only the performance of technologies but also operational constraints that are unique to the Cislunar domain. Models representing both of these factors have been assembled into software packages to enable a Model-Based Systems Engineering (MBSE) analysis of the problem. In order to perform optimization studies across the library of potential models, we have further developed a rapidly configurable Multi-Disciplinary Applied Optimization (MDAO) modeling framework. The MDAO framework uses object-oriented programming techniques to standardize model interfaces and allow them to be integrated into a unified optimization environment that is extended from NASAs OpenMDAO package. At the optimization stage, this MDAO system leverages genetic algorithms to produce the optimal choice of technologies and designs with respect to desired operational performance metrics. The end result is a modular software package that can be used to execute optimizations across a range of present and future cislunar technologies and designs.
We provide an overview of how the metrics for sensor detection performance and orbit maintenance costs can be integrated into a library of optimization metrics. These metrics were used to assess the performance of distributed sensor networks consisting of observers spanning the cislunar space. Sensor detection performance is quantified by the percentage of the coverage over time, whereas the maintenance costs are quantified by orbit instability and navigation uncertainty. These metrics are used to optimize the SDA architecture while considering the operational challenges of deploying and maintaining distributed sensors within the chaotic cislunar volume. We present our results in the form of a Pareto front of non-dominated solutions to each of the specified performance/cost metrics and changes in architecture optimization results depending on which metrics are considered.
Date of Conference: September 19-22, 2023
Track: Cislunar SDA