Shawn W. Abernethy, Jr., Stratagem Group; Emily Gerber, L3Harris; William DeLude, Stratagem Group; Weston Faber, L3Harris; Thomas Kelecy, L3Harris; Taylor Nave, L3Harris
Keywords: Unscented Schmidt-Kalman Filter, Measurement Fusion, constrained optimization, Monte Carlo , Genetic Algorithm, Monte Carlo Markov Chain
Abstract:
Lumos is a suite of software-based capabilities that enable more agile space-based surveillance operations. It reduces the labor involved in SDA anomaly detection and sensor tasking. A near real-time data anomaly filtering process, along with a coordinated multi-constellation optimized tasking scheme, are being established to automate processes for ground operators. The Lumos data anomaly alerts will be provided to a ground operator in near real-time, and the coordinated tasking will provide a “recommendation” for optimal combined collections that minimizes space object uncertainty. This paper will walk through the problem framework and provide example use cases by leveraging simulated data with notional SDA satellite sensors. With this dataset, Lumos efficiently detects a data anomaly from degraded sensor state information by employing a USKF, and it recommends effective coordinated collection using either a Genetic Algorithm or Monte Carlo Markov Chain to maintain custody of an RSO. The operational architecture and Concept of Operations (CONOPS) for Lumos will also be described.
Date of Conference: September 14-17, 2021
Track: SSA/SDA