Ship Collision Avoidance Using Scenario-Based Model Predictive Control**This work was supported by the Research Council of Norway, Statoil, DNV GL and Sintef through the Centers of Excellence funding scheme, Project number 223254 - Centre for Autonomous Marine Operations and Systems (NTNU-AMOS), and the Research Council of Norway, DNV GL, Kongsberg Maritime and Maritime Robotics through the MAROFF knowledge project 244116 Sensor Fusion and Collision Avoidance for Autonomous Surface Vessels.

Abstract

A set of alternative collision avoidance control behaviors are parameterized by two parameters: Offsets to the guidance course angle commanded to the autopilot, and changes to the propulsion command ranging from nominal speed to full reverse. Using predictions of the trajectories of the obstacles and ship, the compliance with the COLREGS rules and collision hazards associated with the alternative control behaviors are evaluated on a finite prediction horizon. The optimal control behavior is computed in a model predictive control implementation strategy. Uncertainty can be accounted for by increasing safety margins or evaluating multiple scenarios for each control behavior. Simulations illustrate the effectiveness in test cases involving multiple dynamic obstacles and uncertainty associated with sensors and predictions.