UUV Path Planning for Collision Avoidance Based on Ant Colony Algorithm

摘要

In this paper, the ant colony algorithm used to study the path planning for collision avoidance of unmanned underwater vehicle (UUV) in a large-scale marine environment. Multi-beam forward-looking sonar's simulation detection model based on the grid environment model established in it. In order to overcome the limitation that the ant colony can only choose to walk on adjacent grids, the concept of virtual visualization introduced in the last step of the ant colony. This paper uses a non-linear penalty mechanism in the objective function of ant colony optimization to avoid large-angle turning movements in online path decision making. Simulation results show that the proposed method has short planning time, high optimization efficiency, and smooth paths.