Embedded State Estimation for Optimization of Cislunar Space Domain Awareness Constellation Design

TL;DR

This paper develops an optimization framework for designing cislunar space-based observation constellations that enhance space domain awareness by improving target state estimation amidst observational challenges.

Contribution

It introduces a novel multi-observer placement optimization method that integrates state estimation performance, tailored for cislunar space SDA challenges.

Findings

Optimized constellations improve state estimation accuracy across various orbit types.

Sensor tasking and fidelity significantly impact constellation performance.

Validated effectiveness against a broad set of target orbits.

Abstract

The traffic in cislunar space is expected to increase over the coming years, leading to a higher likelihood of conjunction events among active satellites, orbital debris, and non-cooperative satellites. This increase necessitates enhanced space domain awareness (SDA) capabilities that include state estimation for targets of interest. Both Earth surface-based and space-based observation platforms in geosynchronous orbit or below face challenges such as range, exclusion, and occlusion that hinder observation. Motivated by the need to place space-based observers in the cislunar space regime to overcome these challenges, this paper proposes a cislunar SDA constellation design and analysis framework that integrates state estimation into an optimization problem for determining the placement of observers for optimal state estimation performance on a set of targets. The proposed multi-observer placement optimization problem samples from a range of possible target orbits. Upon convergence, the optimized constellation is validated against a broader set of targets to assess its effectiveness. Two comparative analyses are presented to evaluate the effects of changes in the sensor tasking procedure and sensor fidelity on the optimized constellation, comparing these to a single observer baseline case. The results demonstrate that the optimized constellations can provide accurate state estimation for various orbit families.