Obstacle avoidance strategies for autonomous surface vehicles

Autonomous Surface Vehicles (ASVs) rely on two layers of guidance for effective path planning: global and local methods. Global methods require a macroscopic view of the mission's goals in order to efficiently navigate an environment. For this research, different global path planner algorithms are explored both in simulation and in field tests to observe path planner qualities for modularity, practicality, and robustness for future integration with local, reactive obstacle avoidance. Three obstacle configurations are tested to study situational approaches for each algorithm. With appropriate inputs for path planning parameters (i.e. connection distance, number of nodes, and iterations), numerical simulations are able to find a feasible path. Results demonstrate path following robustness in response to a generic ASV craft with only a roughly tuned heading and speed controller.