Metraipolis › archive › dc_d524bd09b54d
Markovian Dynamics Enforcer: Feasibility Preserving Correction on Learned Dynamics Manifolds — arXiv:2609.39888
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.39888
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-07T13:46:46.912Z
- SHA-256 of the archived text
2930c294712151b43330602354d16384630c0bca712a27ac2ad56b402cea4e98- Stable URL of this page
https://metraipolis.com/documento/dc_d524bd09b54d
How to cite
Markovian Dynamics Enforcer: Feasibility Preserving Correction on Learned Dynamics Manifolds — arXiv:2609.39888 — arXiv (en), https://arxiv.org/abs/2609.39888 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_d524bd09b54d (Metraipolis archive dc_d524bd09b54d, sha256 2930c294712151b43330602354d16384630c0bca712a27ac2ad56b402cea4e98)
BibTeX
@misc{dcd524bd09b54d,
title = {Markovian Dynamics Enforcer: Feasibility Preserving Correction on Learned Dynamics Manifolds — arXiv:2609.39888},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.39888},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_d524bd09b54d; sha256 2930c294712151b43330602354d16384630c0bca712a27ac2ad56b402cea4e98},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.39888. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Neural trajectory predictors can reach low prediction error while violating dynamics, actuator limits, or state constraints, especially when controls are unobserved and dynamics are partially specified. We introduce the Markovian Dynamics Enforcer (MaDE), a time-invariant post-hoc operator mapping state-transition proposals onto a learned feasible dynamics manifold, trained on feasible states without ground-truth controls. For each transition it infers a control and recomputes the state through a completion model of known physics plus a learned residual. It then corrects that control by gradient-based inequality reduction, so inequality satisfaction is best-effort within an iteration budget. Since every correction iterate re-enters the completion model, the returned state is dynamically consistent by construction relative to that model and the supplied previous-state anchor. MaDE drives dynamics residuals to essentially zero on fully specified simulated systems, and on an underspecified system leaves a smaller true-dynamics residual than the baselines. Designed to attach to arbitrary predictors, the frozen operator is evaluated downstream of recurrent, structured state-space, and transformer predictors. On recorded vehicle trajectories the one-step residual against a kinematic bicycle model is 0.0071 to 0.0072 for MaDE and 0.1703 to 0.1714 for raw predictors. MaDE raises average displacement error by a factor of 1.57 to 1.83. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-10-01 in cs.LG). Categorie arXiv: Machine Learning (cs.LG) ; Robotics (cs.RO); Systems and Control (eess.SY). Pagina del lavoro: https://arxiv.org/abs/2609.39888 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-07T13:18:10.485Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-10-01?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).