Terminal-Aware Hybrid A* for Vehicle Path Planning: Analytic Connections, Coarse Guidance, and a Controlled Study of Adaptive Rollouts
Machines · 14.11.2026 · 14 (11) · 1347
DOI: 10.3390/machines14111347 ↗Kurz gefasst
Fachartikel in der Zeitschrift Machines, online erschienen am 14.11.2026. Verfasst von Xueyan Cheng und 1 weiteren Autor:innen. Die Arbeit nennt 35 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Exact terminal orientation can make forward vehicle Hybrid A* planning harder than positional arrival. This study evaluates an implementation combining Dubins guidance, a coarse obstacle-aware field, analytic terminal connections, and batched collision checking. Adaptive rollouts and delayed bidirectional search are evaluated options; earlier component selection favors a forward fixed-length default. In the original 24 paired tasks on 12 layouts, all seven methods succeed; the default reduces mean runtime by 45.7% and mean sampled length by 5.53% relative to the earlier fixed-3 m engine. The descriptive median paired length reduction is only 0.36%, and three cases contribute 89.02% of the net length saving. The primary post-review analysis adds 933 runs, including matched component ablations and 36 new layouts. The default succeeds on 35/36 new layouts versus 34/36 for the earlier fixed engine. Ablations support a role for coarse guidance and batching but show no all-family advantage on the original tasks. A conservative geometric bound certifies continuous clearance of all 24 original enhanced paths and all 35 successful enhanced paths on the expanded set within the static three-circle model. These results support scoped implementation improvements while leaving reverse driving, dynamic feasibility, uncertain sensing, and embedded execution unvalidated.
Zitieren
APA
Cheng, X., & Dai, L. (2026). Terminal-Aware Hybrid A* for Vehicle Path Planning: Analytic Connections, Coarse Guidance, and a Controlled Study of Adaptive Rollouts. Machines, 14(11), 1347. https://doi.org/10.3390/machines14111347
Vancouver
Cheng X, Dai L. Terminal-Aware Hybrid A* for Vehicle Path Planning: Analytic Connections, Coarse Guidance, and a Controlled Study of Adaptive Rollouts. Machines. 2026;14(11):1347. doi:10.3390/machines14111347
BibTeX
@article{Cheng2026,
author = {Cheng, Xueyan and Dai, Liang},
title = {Terminal-Aware Hybrid A* for Vehicle Path Planning: Analytic Connections, Coarse Guidance, and a Controlled Study of Adaptive Rollouts},
journal = {Machines},
year = {2026},
volume = {14},
number = {11},
pages = {1347},
doi = {10.3390/machines14111347}
}
EUVIMED