A graph-based symbolic Bayesian framework to predict chaotic transport under uncertainty
Chaos: An Interdisciplinary Journal of Nonlinear Science · 01.11.2026 · 36 (11)
DOI: 10.1063/5.0358898 ↗Kurz gefasst
Fachartikel in der Zeitschrift Chaos: An Interdisciplinary Journal of Nonlinear Science, online erschienen am 17.11.2026. Verfasst von David Canales und 1 weiteren Autor:innen. Beteiligte Einrichtung: Department of Aerospace Engineering, Embry-Riddle Aeronautical University, 1 Aerospace Boulevard, Daytona Beach, Florida 32114, USA. Die Arbeit nennt 52 Literaturangaben. Gefördert von: National Science Foundation.
Abstract (Originalsprache)
Predicting long-term transport in nonlinear chaotic dynamical systems is challenging because small initial-state uncertainties may produce qualitatively different future pathways. Conventional state estimators quantify uncertainty in the evolving state but do not explicitly distinguish among qualitatively different transport histories, especially when measurements are sparse. This work develops a graph-based symbolic Bayesian framework for predicting chaotic transport under uncertainty and demonstrates it for the Earth–Moon system in the circular restricted three-body problem. The configuration space is partitioned into dynamically interpretable regions using zero-velocity geometry, libration-point structure, and an isolating-block-motivated lunar boundary. These regions define an energy-dependent directed graph, and trajectories are encoded as symbolic region-transition words. An initial-state estimate and covariance generate a Sobol ensemble. Propagated samples are grouped into symbolic transport classes, defining prior probabilities over competing future histories. Sequential measurements update these probabilities through Bayes’s theorem and progressively concentrate the posterior on the most measurement-consistent pathways. Two numerical cases involving lunar entries through the L2 and L1 gateways show that initially competing transport outcomes are distinguished using sequential position measurements. For the dominant symbolic class, measurement-conditioned library representatives yield position- and velocity-error norms comparable in scale to those of extended Kalman, unscented Kalman, and particle filters at the measurement epochs. The framework, therefore, converts continuous-state uncertainty into a measurement-conditioned probability distribution over dynamically distinct future transport pathways. It provides an interpretable complement to conventional continuous-state estimation for long-term prediction in chaotic transport regimes.
Autor:innen und Institutionen
- Department of Aerospace Engineering, Embry-Riddle Aeronautical University, 1 Aerospace Boulevard, Daytona Beach, Florida 32114, USA.
ORCID-Profile: David Canales, Seur Gi Jo
Förderung
- National Science Foundation
Zitieren
APA
Canales, D., & Jo, S. G. (2026). A graph-based symbolic Bayesian framework to predict chaotic transport under uncertainty. Chaos: An Interdisciplinary Journal of Nonlinear Science, 36(11). https://doi.org/10.1063/5.0358898
Vancouver
Canales D, Jo SG. A graph-based symbolic Bayesian framework to predict chaotic transport under uncertainty. Chaos: An Interdisciplinary Journal of Nonlinear Science. 2026;36(11). doi:10.1063/5.0358898
BibTeX
@article{Canales2026,
author = {Canales, David and Jo, Seur Gi},
title = {A graph-based symbolic Bayesian framework to predict chaotic transport under uncertainty},
journal = {Chaos: An Interdisciplinary Journal of Nonlinear Science},
year = {2026},
volume = {36},
number = {11},
doi = {10.1063/5.0358898}
}
EUVIMED