Research On Path Planning Of Unmanned Surface Vehicle Patrol

Maritime patrol is an important area for the intelligent application of unmanned surface vehicles,which can achieve all-weather operations,save labors and reduce the risks of maritime operations.Now it has become a research hotspot in various countries around the world.Firstly,in order to achieve fully autonomous navigation,it is necessary for unmanned surface vehicles to solve the problem of path planning,where it is a basic requirement for patrol path planning to traverse each patrol point quickly and accurately.Moreover,with the emergence of unknown obstacles during navigation,it is indispensable to solve the problem of dynamic collision avoidance.Therefore,in order to improve the economy and safety of autonomous patrol of unmanned surface vehicles,the paper conducts in-depth research on the path planning of unmanned surface vehicles.The main research work includes the following points:First,the paper designed an A* algorithm with multi-directional search,and used it to build the shortest path network between various patrol points of unmanned surface vehicles.Aiming at the surface navigation environment,an environmental model was established by using the grid method,in which the grid size refers to the size of the unmanned surface vehicles.Based on the grid modeling,the traditional eight-direction search A* algorithm was improved in multiple directions.The principle and implementation steps of the improved search algorithm were analyzed in detail.The improved A* algorithm was used to search the shortest path between each two patrol points,and a patrol path network was constructed based on the shortest path between any two points.Secondly,the paper designed a smoothing ant colony algorithm to solve the shortest path of patrol.Based on the shortest distance between patrol points of the patrol path network,a distance matrix was established to transform the patrol problem into a completely undirected classic traveling salesman problem.The mathematical model of the patrol optimization problem was established with the shortest patrol distance as the target and the patrol point not being repeated as a constraint.The ant colony algorithm was used to design the solution method of patrol optimization problem.Considering the limitation of the dynamic characteristics of unmanned surface vehicles,it cannot be deflected at a large angle.A smoothing algorithm was added based on the ant colony algorithm to solve this problem,and a smooth arc curve was used instead of the turning path to smooth each turning point.The patrol path smoothing simulation analysis was performed in the actual environment.The results showed that the smoothed patrol path not only greatly reduced the total turning angle,but also fit more the actual navigation requirements of unmanned surface vehicles.Finally,in view of the collision avoidance of unknown obstacles,this paper designed a local path planning method combining dynamic window method and velocity obstacle method.Aiming at the problem that the velocity information of unmanned surface vehicles and obstacles were not considered in the dynamic window method,the velocity vectors in the velocity set of the dynamic window method were filtered according to the principle of the velocity obstacle method,and the velocity vectors that did not collide with obstacles were retained.The simulation analysis in the actual environment showed that the improved local path planning algorithm was safer and more reliable than the traditional dynamic window method for obstacle avoidance.In summary,the paper carried out static path planning and dynamic path planning for patrol problems.The simulation results showed that the designed patrol path planning method provided a feasible solution for unmanned surface vehicles autonomous patrol.