Improved Multi-Scale A* and DWA Algorithm for USV Path Planning Integrating MMG Maneuvering Constraints
Journal of Marine Science and Engineering · 27.07.2026 · 14 (15) · 1369
DOI: 10.3390/jmse14151369 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Marine Science and Engineering, online erschienen am 27.07.2026. Verfasst von Huiyun Xue und 2 weiteren Autor:innen. Die Arbeit nennt 27 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Applied Basic Research Program in the Field of Artificial Intelligence under the Liaoning Province Department of Science and Technology, Industry-University-Research Collaborative Education Project of Liaoning Province, High-Level Talent Innovation Program of Dalian.
Abstract (Originalsprache)
To balance safety, real-time performance, and dynamic feasibility in path planning for unmanned surface vehicles (USVs) operating in complex aquatic environments, the study proposes a hybrid planning framework that integrates a multi-scale A* algorithm with the Dynamic Window Approach (DWA) under Maneuvering Modeling Group (MMG) dynamic constraints. Global path optimization incorporates target-oriented neighborhood expansion, obstacle risk constraints, and a maneuverability cost function to comprehensively evaluate path length and collision risk, thereby enhancing the efficiency of the hybrid global-local planning framework. Local trajectory prediction strictly couples the MMG USV motion model with the DWA algorithm. Real-time optimization of these dynamically constrained trajectories ultimately achieves proactive and safe evasion of dynamic obstacles. Compared with the traditional A* + DWA method, the proposed hybrid method reduces path length and computational time by 5.09% and 31.93%, effectively suppressing abrupt heading changes during dynamic obstacle avoidance. This paper provides an efficient path planning strategy that conforms to USV maneuvering constraints for autonomous navigation and dynamic obstacle avoidance.
Förderung
- Applied Basic Research Program in the Field of Artificial Intelligence under the Liaoning Province Department of Science and Technology
- Industry-University-Research Collaborative Education Project of Liaoning Province
- High-Level Talent Innovation Program of Dalian
- Scientific Research Startup Fund for Introduced PhD Talents
Zitieren
APA
Xue, H., Song, C., & Sui, J. (2026). Improved Multi-Scale A* and DWA Algorithm for USV Path Planning Integrating MMG Maneuvering Constraints. Journal of Marine Science and Engineering, 14(15), 1369. https://doi.org/10.3390/jmse14151369
Vancouver
Xue H, Song C, Sui J. Improved Multi-Scale A* and DWA Algorithm for USV Path Planning Integrating MMG Maneuvering Constraints. Journal of Marine Science and Engineering. 2026;14(15):1369. doi:10.3390/jmse14151369
BibTeX
@article{Xue2026,
author = {Xue, Huiyun and Song, Chunyu and Sui, Jianghua},
title = {Improved Multi-Scale A* and DWA Algorithm for USV Path Planning Integrating MMG Maneuvering Constraints},
journal = {Journal of Marine Science and Engineering},
year = {2026},
volume = {14},
number = {15},
pages = {1369},
doi = {10.3390/jmse14151369}
}
EUVIMED