Uncertainty Propagation of Correlated Quaternion and Euclidean States using Partially-Conditioned Gaussian Mixtures


The partially-conditioned Gaussian (PCG) density, a variant of the Gauss-Bingham density, quantifies the uncertainty of a state vector comprised of an attitude quaternion and other Euclidean states on their natural manifold, the unit hypercylinder. The conditioned Gaussian density is first developed by conditioning a Gaussian density on the unit hypersphere, and is an alternate representation of the Bingham density. The PCG density is then developed, which conditions only the quaternion portion of the aforementioned state vector on the unit hypersphere. The PCG density is then extended to the PCG mixture density, which can be used to approximate an arbitrary density on the unit hypercylinder. A method to construct a PCG mixture density approximating the PCG density on the two-dimensional cylinder is then developed. The temporal evolution of the PCG mixture density given system dynamics is then quantified and is compared to a Monte Carlo approach in order to verify its performance.

Meeting Name

19th International Conference on Information Fusion (2016: Jul. 5-8, Heidelberg, GE)


Mechanical and Aerospace Engineering

Keywords and Phrases

Bins; Information fusion; Gaussian density; Gaussian mixtures; Mixture density; Monte Carlo approach; System Dynamics; Temporal evolution; Two-dimensional cylinder; Uncertainty propagation; Gaussian distribution

International Standard Book Number (ISBN)


Document Type

Article - Conference proceedings

Document Version


File Type





© 2016 Institute of Electrical and Electronics Engineers (IEEE), All rights reserved.

Publication Date

01 Aug 2016

This document is currently not available here.