Research On Algorithms Of Direct Navigation And Patrol Path Planning For Unmanned Vehicle

With the rise of unmanned equipment,the research and application of surface unmanned vehicles have gradually received attention.The unmanned vessel path planning technology is an important subject in the research of unmanned vessel platforms.It is an important part of the realization of the intelligent unmanned driving function,and it is also one of the key technologies to ensure the safe navigation of unmanned vessel.In order to obtain the shortest navigation path to ensure the navigation economy of unmanned vessel,the thesis carries out in-depth research on the shortest path planning problem of direct navigation and multi-point patrol for unmanned vessel.The main research work is as follows:Firstly,the thesis summarizes the development and current situation of surface unmanned vessel and their path planning algorithms,and analyzes the path planning problem of surface unmanned vessel.The thesis summarizes the common navigation environment modeling methods and the path search method,and analyzes the applicable occasions and performance characteristics of various path planning methods,which provides a theoretical basis for the subsequent research of the thesis.Secondly,a shortest path search method between two navigation points in the water surface feasible domain based on the Dijkstra algorithm is designed.Aiming at the navigation environment of unmanned vessel,the environment model by grid method is established to distinguish obstacles and feasible fields.On the basis of grid map,the basic Dijkstra algorithm is improved and three path planning algorithms include four-direction search,eight-direction search and multi-angle search are designed in the thesis.The principle of various search algorithms is deduced in detail,and the specific implementation methods are given.The performance characteristics of various algorithms are analyzed.The designed path planning method is tested by the actual map.The test results show that the improved path search algorithm obtains a shorter path than the basic Dijkstra algorithm.Thirdly,a shortest path planning method for multi-point patrol is designed.This method uses the optimization idea to transform the multi-point patrol shortest path planning problem into a traveling salesman problem.The above proposed shortest path search algorithm of the feasible field navigation is used to obtain the shortest path connectivity network map between the patrol points,and a mathematical description of the patrol path optimization problem is established.Aiming at the patrol path optimization problem,a path optimization algorithm based on genetic algorithm is designed,and the mutation operator of genetic algorithm is improved according to the greedy principle.The designed unmanned vessel patrol path planning method is tested by actual map.The test results show that the designed patrol path optimization algorithm can obtain the shortest global patrol path,and the improved genetic algorithm used for path search has a faster convergence speed than traditional genetic algorithm.In summary,the designed path planning method of feasible field navigation and patrol for the unmanned vessel by the thesis has initially solved the path planning requirements for unmanned vessels during autonomous navigation,providing basic guarantee for safe,economical and fast navigation of unmanned vessels.