Multiple Hypothesis Tracking Filter with Lambert Post Filtering for Multiple Space Object Tracking
Abstract
The data association problem occurs when there is uncertainty and ambiguity in assessing the origin of measurements especially when the measurements are being made from multiple locations at multiple times. For example, in the Space Situational Awareness (SSA) problem, of tracking catalogued Resident Space Objects, it is important to discern the objects being tracked from other debris, non-catalogued objects, maneuvering satellites and even random or spurious measurements. In situations with multiple targets in the field of view, sensor generates multiple measurements where the origin of each measurement is not readily evident. The target to track association problem thus forms an integral preprocessing step for catalog maintenance and uncertainty quantification of the space objects of interest. Most filter implementations assume that the measurements to be of the same physical target being tracked. This scheme can lead to wrong estimates and lost tracks due to incorrect target-measurement assignments. A Multi Hypothesis Tracking (MHT) filter, which combinatorically performs feasible target-measurement assignments is implemented to solve the data association problem. The MHT algorithm, has been augmented with Lambert problem based Initial Orbit Determination (IOD) technique. IOD has been used for target initializations and also as a post filtering sanity check, on batches of measurements which have been assigned to a target, in order to arrest the combinatorial growth of the possible assignment hypothesis.
Subject
Multi Target trackingJoint Probabilistic Data Association
Filtering
Estimation
Space Situational Awareness
Orbital debris
Tracking
Astrodynamics
Citation
Mishra, Utkarsh Ranjan (2019). Multiple Hypothesis Tracking Filter with Lambert Post Filtering for Multiple Space Object Tracking. Master's thesis, Texas A&M University. Available electronically from https : / /hdl .handle .net /1969 .1 /187938.