Efficient simulation of wave glider dynamics and validation with trans-Pacific voyages

OCEANS 2023 - Limerick(2023)

引用 1|浏览2
暂无评分
摘要
The decentralised control architecture of robot swarms may give them the robustness required for prolonged operations in hostile environments such as the maritime sector. A swarm of autonomous marine vehicles could form dynamic, spatially reconfigurable sensor networks for monitoring marine phenomena. In this paper, we present methods for simulating the dynamics of a wave glider, an autonomous surface vehicle with high endurance, using the ASVLite simulator. As the wave glider’s speed depends on the encountered waves and ocean currents, our method replicates ocean waves and currents based on archives of open-source weather data. The model is calibrated using recorded onboard data from three wave gliders involved in a trans-Pacific voyage lasting over 16 months, including sea states up to 7 m of significant wave height. We conducted simulations of 10,325 segments from a fourth wave glider’s voyage to validate our calibrated model. The findings demonstrate that our approach effectively captures both the magnitude and variation in the wave glider’s speeds as it responds to the encountered wave heights and ocean currents. To demonstrate the application of our simulator for planning environmental monitoring missions, we also simulate and compare a selected voyage segment where one of the wave gliders sails through a category-4 tropical cyclone in the South Pacific. For the selected segment, the median of the cumulative position error—the difference between simulation and reality—of the glider was 14.68 km over the total length of the journey of 370 km. Similarly, the maximum cumulative time errors measured at waypoints along the glider’s path was within ±10 hrs over the total duration of 147 hrs. Our results demonstrate that our simulation method using the ASVLite simulator and open-source wave and current data exhibits good performance and fidelity, making it a potential tool for planning long-range environmental monitoring missions using wave gliders.
更多
查看译文
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要