COLREGs-Informed RRT* for Collision Avoidance of Marine Crafts

Abstract

The paper proposes novel sampling strategies to compute the optimal path alteration of a surface vessel sailing in close quarters. Such strategy directly encodes the rules for safe navigation at sea, by exploiting the concept of minimal ship domain to determine the compliant region where the path deviation is to be generated. The sampling strategy is integrated within the optimal rapidly-exploring random tree algorithm, which minimizes the length of the path deviation. Further, the feasibility of the path with respect to the steering characteristics of own ship is verified by ensuring that the position of the new waypoints respects the minimum turning radius of the vessel. The proposed sampling strategy brings a significant performance improvement both in terms of optimal cost, computational speed and convergence rate.

Recent advances in Rapidly-exploring random tree: A review

“…In Ref. [ 99 ], the design of the short-horizon planner for collision avoidance is addressed, and an improved Informed-RRT* algorithm is proposed, which encodes the International Regulations for Preventing Collisions at Sea (COLREGs) by defining a rule-compliant region in the configuration space and searching for feasible paths within that region.…”

Safe Reinforcement Learning for Urban Driving using Invariably Safe Braking Sets

“…b) Motion planning for vessels: Maritime motion planners can be divided into three building blocks [30]: a guidance system generating reference trajectories, a control system for tracking reference trajectories, and a state observer 1 . For example, one line of research employs search-based algorithms based on motion primitives, e.g., rapidly-exploring random trees [28], [31], [32]. Other studies employ model predictive control (MPC) [25], [27], [33] to obtain an optimal control signal.…”

A Dynamic Programming Approach to Collision Avoidance of Autonomous Ships

“…This requirement is essential in scenarios where autonomous systems interact with human-controlled systems [13]. The scientific literature proposed various approaches to collision avoidance of marine vessels, including A* [14,15], Dijkstra's algorithm [16,17], visibility graphs [18], rapidly-exploring random trees [19][20][21], and various population-based heuristics [22][23][24][25].…”