Research On Autonomous Collision Avoidance Method Of Unmanned Surface Vehicles

More than two-thirds of the earth is covered by oceans,which contains rich resources.Countries increasingly pay more and more attention to the exploitation and utilization of marine resources.As a new surface unmanned system platform,USV has been widely concerned and applied in many fields,such as environmental monitoring,marine investigation,security rescue.The USV navigates based on the global planning path and senses the dynamic changes of the surrounding environment in real time through the navigation radar or other types of sensors.It needs to dynamically adjust the track points to avoid obstacles autonomously according to the perceived surrounding state information.For the situation that obstacles approach USV quickly and closely in the unknown environment,this thesis has conducted in-depth research.The main research contents are as follows:1)Aiming at the problem of Near-field reactive collision avoidance of USV in unknown environment,two typical collision avoidance algorithms of traditional DWA and traditional APF algorithm are studied,and an improved DWA algorithm is designed.By predicting the trajectory of obstacles,deleting the speed at which collision is expected to occur,and modifying the distance evaluation function,the problem of poor effect of avoiding dynamic obstacles is solved.In addition,QPSO is used to find the optimal speed in the dynamic speed space set and the rule evaluation function is introduced to improve the degree of compliance with the rules in the collision avoidance process of USV,and improve the stability,comfort and safety in the navigation process.2)Aiming at the problems existing when the artificial potential field method is applied to the autonomous collision avoidance of USV,an improved APF algorithm is designed by modifying the gravitational and repulsive potential field functions,so that the USV can realize the safe collision avoidance with dynamic obstacle ships under the condition of abiding by the collision avoidance rules.In this thesis,the gravitational force of the established route to the USV is designed,and the USV can return to the previous route after avoiding obstacles ship.By introducing the influence range of repulsive force of dynamic obstacles to determine the timing of repulsive force exerted by obstacles of different speeds on USV,the ability of USV to avoid collision with dynamic obstacles is improved.Combined with the maneuverability constraints of USV,the heading angle is output to improve the smoothness and safety of collision avoidance trajectory.3)The experiment of autonomous collision avoidance system based on small USV was carried out.USV and obstacle ship are selected to test the autonomous collision avoidance system in two typical encounter scenarios to verify the effectiveness of the improved collision avoidance algorithm in this thesis.The experimental results show that the improved collision avoidance algorithm can avoid the obstacle ship efficiently and safely in the face of the sudden obstacle ship.