Autonomous Obstacle Avoidance For Unmanned Surface Vehicles Based On The Hybrid Intelligent Algorithm

With the continuous development of economy and society and the increasing number of people,how to efficiently explore,discover and utilize marine resources has become a new topic for mankind.As a result of the rapid development of artificial intelligence,unmanned surface vehicles,because of its unique role in the marine field,have attracted the attention of scholars from all sides.Among them,autonomous navigation technology and autonomous obstacle avoidance technology have become the core areas of intelligent unmanned surface vehicles,and are also the main technical difficulties facing development.In a complex and ever-changing marine environment,an unmanned surface vehicle must formulate a navigation plan in real time according to the ocean conditions,and at the same time ensure that the obstacle avoidance measures taken comply with the International Regulations for Preventing Collisions at Sea and meet the ship’s motion performance and take into account unknown obstacles.Obstacle avoidance measures that may only be needed by objects.Therefore,this article mainly studies the following aspects of the collision avoidance plan for high-speed unmanned surface vehicles:(1)By consulting a large number of documents at home and abroad and the research on unmanned surface vehicles by relevant agencies,the research status of unmanned surface vehicles and autonomous obstacle avoidance is summarized,and the unmanned surface autonomous obstacle avoidance algorithm is mainly divided into based on Behavioral reactive obstacle avoidance algorithm and local path planning algorithm based on path search.(2)Aiming at the characteristics of high degree of ambiguity of ship collision risk,the fuzzy comprehensive evaluation method is used to establish a fuzzy membership function for the key parameters of collision decision,and the weight of the membership function is selected according to the maritime navigation experience,and the reliability of the model is verified by simulation And security.(3)In order to solve the problem of particle swarm algorithm that is easy to fall into local extremum and the optimization speed is too slow in some cases,the simulated annealing algorithm and particle swarm algorithm are combined to improve it,and the traditional particle swarm algorithm is replaced by Hammersley point set Random initialization makes the particle distribution more uniform and the algorithm performance more stable.At the same time,adaptive inertia weights based on population distribution density are used to replace traditional linear inertia weights,which makes the algorithm’s local search ability and global optimization ability more balanced.(4)Establish an unmanned surface vehicle’s decision model for avoiding obstacles at sea.Through the analysis of the International Regulations for Preventing Collisions at Sea,three types of encounters,namely,encounter,overtake and cross,are divided.Combined with the safety angle and the International Maritime Collision Avoidance Rules,the obstacles are reversed and eccentrically expanded,and an unmanned surface vehicle autonomous obstacle avoidance strategy combining the speed obstacle method and adaptive particle swarm optimization based on simulated annealing algorithm is designed.It will verify the feasibility and robustness of the algorithm under the situation.