A 3D dynamic Voronoi diagram-based path-planning system for UUVs

Abstract

This paper proposes a rapid path-planning and replanning system for Unmanned Underwater Vehicles (UUVs) that navigate in environments where subsea structures and other vehicles may be present. The proposed method is based on the Voronoi diagram, which is used to generate an initial set of connected waypoints (a roadmap) in the three-dimensional (3D) space, ensuring a certain clearance to avoid collisions with obstacles or grounding (e.g. collision with the ground). A 3D continuous path, composed by straight segments and circumference arcs, connects the aforementioned waypoints. If the vehicle encounters any moving or static obstacle and a collision risk is detected, the path is replanned online. In this context, an evaluation of the risk must be performed, and a list of traffic rules inspired by the International Regulations for Preventing Collisions at Sea (COLREGs, which are adopted for surface vessels), and by the Traffic and Collision Avoidance System (TCAS, which is adopted for aerial vehicles), is proposed for underwater vehicles. Those rules define the replanning procedure, so that an effective and safe collision avoidance maneuver can be performed whenever necessary. Simulations are performed on an subsea factory scenario, and results are presented to show the effectiveness of the proposed method.