Ship intelligence has been a hot topic in recent years. How to achieve autonomous maneuvers in a complex marine environment in a safe, efficient, and low-cost manner is a fundamental task that ocean engineers face. This paper presents a two-stage trajectory planning scheme to address the minimum-time maneuvering problem in close-range encounters. The scheme is robust and versatile, as it can deal with the complex spatial variability, such as sea current, state constraints, marine traffic, and physical constraints, of close-range maneuvering. In the first stage, a directed graph with variable length is generated according to the sea current distribution. A wavefront search is applied on the graph to explore the reachability, the cost of state constraints, and the risk of collision. After a discrete solution has been found, the second stage involves searching for a smooth solution. A Bézier curve based parameter optimization approach is proposed to get rid of limited moving directions in the directed graph and explore around the discrete path. The result will be a near-optimal, smooth path. The proposed scheme has been tested to solve the Zermelo's ship steering problem and several other close-range maneuvering problems. The results demonstrate that the scheme is efficient in generating smoothed minimum-time trajectories for surface vessels when maneuvering in close-range encounters.

中文翻译：

---

近距离接触自主船舶操纵的时间最优轨迹规划

船舶智能是近年来的热门话题。如何在复杂的海洋环境中以安全、高效、低成本的方式实现自主机动，是海洋工程师面临的一项基本任务。本文提出了一种两阶段轨迹规划方案，以解决近距离遭遇中的最小时间机动问题。该方案稳健且通用，因为它可以处理复杂的空间变化，例如近距离机动的海流、状态约束、海上交通和物理约束。在第一阶段，根据海流分布生成可变长度的有向图。在图上应用波前搜索来探索可达性、状态约束的成本和碰撞风险。找到离散解后，第二阶段涉及寻找平滑的解决方案。提出了一种基于贝塞尔曲线的参数优化方法，以摆脱有向图中的有限移动方向并围绕离散路径进行探索。结果将是一条近乎最优的平滑路径。所提出的方案已经过测试，可以解决 Zermelo 的船舶转向问题和其他几个近距离机动问题。结果表明，该方案在近距离遭遇机动时为水面舰艇生成平滑的最小时间轨迹是有效的。所提出的方案已经过测试，可以解决 Zermelo 的船舶转向问题和其他几个近距离机动问题。结果表明，该方案在近距离遭遇机动时为水面舰艇生成平滑的最小时间轨迹是有效的。所提出的方案已经过测试，可以解决 Zermelo 的船舶转向问题和其他几个近距离机动问题。结果表明，该方案在近距离遭遇机动时为水面舰艇生成平滑的最小时间轨迹是有效的。