Target Birth Modeling, Data Fusion, and Robust Filtering for Space Situational Awareness

Abstract

The improved monitoring of satellites in the GEO zone, and their protection from potential collisions with space debris, is of primary concern. Therefore, this proposal suggests the collaborative use of multiple low resolution, low cost telescopes at different locations, within the footprints of geosynchronous satellites and their neighborhoods, offering the potential for greater observations and improved cataloged data in such vicinities. Recently, Random Finite Set (RFS) based filters have been applied to tracking resident space objects (RSOs) in low cost, telescopic images. This proposal therefore advocates research to improve the performance of RFS algorithms in three distinct ways. Firstly, during their track prediction stage, RFS based filters require birth models, which allow new RSO tracks to be initialized or previously mis-detected tracks to be recovered.State of the art RFS trackers implement such models in either a random fashion, to allow tracks to start in orbits beyond current estimated tracks, or based upon inverse measurement models, allowing tracks to be recovered from possible past missed detections. This proposal suggests that previous work on probabilistic admissible regions (PARs) is better suited to generate target births within an RFS based tracker. Secondly, the fusion of multiple, low cost telescopic images is suggested for the longer time observation and increased accuracy of multi-target tracks within the GEO zone. This is intended to increase the tracking performance of potential RSO threats near to satellites. State of the art RFS based trackers are sensitive to the choice of detection parameters, which are telescope, time and weather dependent, requiring rigorous sensor models. Finally, the development of robust RFS trackers, capable of jointly estimating these parameters with the multi-target state, within the Bayesian recursion, is proposed; this could ease the inclusion of new telescopes into a sensor network, for improved SSA.

Document Details

Document Type
DoD Grant Award
Publication Date
Apr 09, 2018
Source ID
FA95501710386

Entities

People

  • Martin Adams

Organizations

  • Air Force Office of Scientific Research
  • United States Air Force
  • University of Chile

Tags

Readers

  • Adaptive Control and Estimation with Uncertainty in Dynamic Systems.
  • Sensor Fusion and Tracking Systems.
  • Space Exploration and Orbital Mechanics.

Technology Areas

  • AI & ML
  • AI & ML - Machine Learning Algorithms
  • Space
  • Space - Space Objects