A Composite Inverse Problem Approach for the On-Orbit Debris Detection and Orbital Determination Problems

In this work we propose solutions to the debris detection and orbital determination (DD & OD) problems that requires a single on-orbit debris detection satellite in a combined pipeline. We introduce the Star tracker Enhanced Noise-Tolerant Image Extraction and Localization (SENTINEL) algorithm that frames the DD & OD problems as a composite inverse problem for the extraction of debris from noisy images and angles-only orbital determination. The DD problem is approached via sequential filtering, robust background extraction, and tracklet filtration. The aforementioned approach allows for reliable extraction of debris in the low SNR regime with image corruptions. This process renders a sequence of relative angles from an on-orbit observer. The OD approach expands the current state-of-the-art methodology for approximating relative orbital elements (ROE) from the Taylor series expansion of relative orbit dynamics. We demonstrate the performance of our algorithm in a synthetic on-orbit debris detection environment.

Forthcoming