Jesse Greaves, University of Colorado Boulder; Daniel Scheeres, University of Colorado Boulder
Keywords: cislunar, ssa, optical, maneuver, perturbation, event, detection, identification, characterization, estimation, labeling, OCBE, EM
Abstract:
Cislunar space is set to become densely trafficked due to international interest in sustained development of the region. This population boom along with the chaotic dynamics of cislunar travel will result in navigational challenges that demand enhanced observational abilities to manage. To improve cislunar observation the ballistic Optimal Control Based Estimator (OCBE) can be utilized to identify unmodeled perturbations, and subsequently identified events can be statistically categorized using uncertainty quantification methods. This methodology not only reduces the amount of observation time needed to identify maneuvers, but also provides a statistical framework to label the maneuvers. This methodology is simulated with observers at multiple cislunar locations to determine visibility properties, OCBE control policy distributions, and maneuver detection performance. These simulations both demonstrate the visibility challenges associated with optical observation in cislunar space and the advantages of the newly developed method for maneuver detection.
There is a great international interest in developing cislunar space for future exploration and exploitation. First and foremost, NASA’s Moon to Mars program and Lunar Gateway project are exemplary of this growing interest since they plan to develop cislunar space into an outpost and highway for deep space exploration. This vision suggests that there will be a myriad of spacecraft flying through cislunar space; everything from human and cargo transport ships to interplanetary vehicles. However, NASA isn’t the only organization aiming at developing cislunar space. ESA, JAXA, Roscosmos, CNSA, ISRO, and ISA have all publicly expressed interest in cislunar space which will create a significant surge in traffic, and inevitably, debris resulting from that traffic. Congestion on this interplanetary highway will be subject to chaotic dynamics, and therefore will need to be monitored carefully to protect and enable long-term use.
The chaotic dynamics of cislunar travel are well-known and lead to several difficulties in estimating cislunar spacecraft motion. Chaotic dynamics mean that small perturbations and estimation inaccuracies will lead to large predictive errors, which lead to filter divergence if unaccounted for or if not enough information is ingested into the system. These dynamics also mandate a station keeping scheme for all spacecraft, which will need to be separated from other maneuvers to maintain an understanding of system health. If station keeping and other events aren’t identified fast enough then the spacecraft will deviate far enough from the nominal and the track will be lost. This necessitates accurate and active statistical estimation to mitigate.
Traditional high accuracy estimation methods for cislunar spacecraft are based around Earth-based radar measurements like the DSN. While this can provide accurate solutions for cooperative vehicles, this leads to several problems for persistent observation. First, scheduling DSN time is difficult because it is already heavily subscribed. Second, the DSN isn’t optimized for non-cooperative vehicles and therefore cannot provide the same level of accuracy for them. Third, only taking measurement from Earth severely limits viewing geometry which leads to limited observational capacity. In order to alleviate these issues cislunar observation platforms can be employed. By placing observers around cislunar space with optical measurement devices, persistent and accurate observation of cooperative and non-cooperative targets can be maintained.
This paper analyzes several observer geometries and measurements to address the visibility issues presented by cislunar observation. These enabling geometries allows for accurate estimation of cislunar systems but does not account for maneuvering spacecraft. The Optimal Control Based Estimator (OCBE), which provides additional statistics that are sensitive to unmodeled forces, is then used to build a station keeping optimal control policy distribution on training data. By creating an expected distribution of the OCBE statistics, a given system’s movement and health can be tracked to a much finer level. Being able to identify events enables accurate and robust estimation of cislunar craft. Being capable of providing this consistent and robust tracking will reduce the uncertainty for cooperative and non-cooperative systems alike, ensuring the safety of all the systems in cislunar space.
Date of Conference: September 15-18, 2020
Best Paper Award Winner 2020
Track: Cislunar SSA