Research on Obstacle Avoidance Scenario Design for Unmanned Surface Vehicle Based on Q-Learning

_ Regarding the design of test scenarios for evaluating the performance of obstacle avoidance algorithms in unmanned surface vehicles (USVs), this article proposes a test scenario constructing method based on Q-learning algorithm. Considering the radius and layout position of obstacles, with the constraints that obstacles are placed within the boundaries of the test scenario and no overlap exists between obstacles, setting the cost of the obstacle avoidance path as the evaluation target, the placement state of obstacles and the translation update action are designed. A minimum distance translation method is employed to handle design schemes that violate the placement constraints during the iteration process. Through testing and comparison with obstacle avoidance algorithms of typical A* and ant colony optimization, as well as the obstacle layout algorithm of genetic optimization, the results demonstrate that the obstacle layout scheme constructed by the Q-learning algorithm achieves superior performance. Compared with manually designed scenarios, the path costs of the two obstacle avoidance algorithms are increased by 15.8% and 17.3%, respectively, effectively testing the path search complexity of the obstacle avoidance algorithms and providing a reference for the automatic construction of USV obstacle avoidance test scenarios. Keywords test scenario design; unmanned surface vehicle; obstacle avoidance; Q-learning

Authors

Publication Details

Journal
Journal of Ship Production and Design
Published
2026-11-19
DOI
https://doi.org/10.5957/jspd.02260004
Primary Topic
Robotic Path Planning Algorithms
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Research on Obstacle Avoidance Scenario Design for Unmanned Surface Vehicle Based on Q-Learning

Zengwu Liu, Q.J. Zhang, Ling Mao, Jin Lin et al.
Journal of Ship Production and Design
Robotic Path Planning Algorithms
article

Research on Obstacle Avoidance Scenario Design for Unmanned Surface Vehicle Based on Q-Learning

Zengwu Liu, Q.J. Zhang, Ling Mao, Jin Lin, Jinglin Du
article en

Abstract

_ Regarding the design of test scenarios for evaluating the performance of obstacle avoidance algorithms in unmanned surface vehicles (USVs), this article proposes a test scenario constructing method based on Q-learning algorithm. Considering the radius and layout position of obstacles, with the constraints that obstacles are placed within the boundaries of the test scenario and no overlap exists between obstacles, setting the cost of the obstacle avoidance path as the evaluation target, the placement state of obstacles and the translation update action are designed. A minimum distance translation method is employed to handle design schemes that violate the placement constraints during the iteration process. Through testing and comparison with obstacle avoidance algorithms of typical A* and ant colony optimization, as well as the obstacle layout algorithm of genetic optimization, the results demonstrate that the obstacle layout scheme constructed by the Q-learning algorithm achieves superior performance. Compared with manually designed scenarios, the path costs of the two obstacle avoidance algorithms are increased by 15.8% and 17.3%, respectively, effectively testing the path search complexity of the obstacle avoidance algorithms and providing a reference for the automatic construction of USV obstacle avoidance test scenarios. Keywords test scenario design; unmanned surface vehicle; obstacle avoidance; Q-learning

Journal of Ship Production and Design
Robotic Path Planning Algorithms
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon