Back to Search Start Over

Improved Dynamic Window Approach for Unmanned Surface Vehicles’ Local Path Planning Considering the Impact of Environmental Factors

Authors :
Zhenyu Wang
Yan Liang
Changwei Gong
Yichang Zhou
Cen Zeng
Songli Zhu
Source :
Sensors, Vol 22, Iss 14, p 5181 (2022)
Publication Year :
2022
Publisher :
MDPI AG, 2022.

Abstract

The aim of local path planning for unmanned surface vehicles (USVs) is to avoid unknown dynamic or static obstacles. However, current relative studies have not fully considered the impact of ocean environmental factors which significantly increase the control difficulty and collision risk of USVs. Therefore, this work studies two ocean environmental factors, namely, wave and current, given that they both have a significant impact on USVs. Furthermore, we redesign a kinematic model of an USV and the evaluation function of a classical and practical local path planning method based on the dynamic window approach (DWA). As shown by the results of the simulations, the path length was impacted mainly by the intensity of the environmental load and slightly by the direction of the environmental load, but the navigation time was significantly influenced by both. Taking the situation in still water as a benchmark in terms of the intensity and direction of the environmental factors, the maximum change rates of the path length were 8.6% and 0.6%, respectively, but the maximum change rates of the navigating time were 17.9% and 25.6%, separately. In addition, the average calculation time of each cycle was only 0.0418 s, and the longest time did not exceed the simulation time corresponding to a single cycle of 0.1 s. This method has proven to be a good candidate for real-time local path planning of USVs since it systematically considers the impact of waves and currents on the navigation of USVs, and thus ensures that USVs can adjust to the planned path in time and avoid obstacles when navigating in the real ocean environment.

Details

Language :
English
ISSN :
14248220
Volume :
22
Issue :
14
Database :
Directory of Open Access Journals
Journal :
Sensors
Publication Type :
Academic Journal
Accession number :
edsdoj.51dfd1480ea44a1bcd6e0e0370c730c
Document Type :
article
Full Text :
https://doi.org/10.3390/s22145181