Research On The Path Planning And Collision Avoidance Methods Of The USV

With the continuous development of global economy,the population of the world is growing day by day.The direction of human exploration and new energy exploration has gradually shifted from land resources to marine resources.As a major tool for exploring the unknown marine environment,the unmanned surface vehicle has received more and more attention for its autonomous control research and has been moving towards more intelligent,systematic and standardization.The path planning of collision avoidance is the key to autonomous control of unmanned surface vehicle.As the marine environment is complex and changeable,it is also necessary to keep in mind the interference of the offshore wind waves on unmanned surface watercraft.Therefore,before an unmanned surface vehicle enters the water area,a global path will be planned based on the known relevant marine environment information.After entering the water area,the unmanned surface vehicle will navigate according to the planned route.And because of the unpredictable conditions(sudden ships,drifting objects,etc.)that may occur in the process of sailing,the unmanned surface vehicle must also detect the surrounding conditions from time to time to obtain the surrounding environment information on these emergencies,the surface unmanned vehicles should be able to quickly detect and respond to different situations to make an accurate and rapid response to adjust their own navigation status,and ultimately avoid obstacles,according to mission requirements to reach the specified goal point.In this paper,the research on the method of avoidance of the path of unmanned surface vehicle is mainly from two aspects.A genetic algorithm based on the static global path planning GA,and the other is the combination of genetic algorithm and dynamic window dynamic collision avoidance.Firstly,genetic algorithm is used to carry out the global path planning of unmanned surface vehicle.Firstly,the heuristic method is used to initialize the population.Based on the actual situation of unmanned sea voyage,the safety,smoothness and economy are added as the evaluation factors to the adaptive Degree function,we introduce delete operation,repair operation and smoothing operation to update the population in the basic operators of genetic algorithm(selection,crossover,mutation),so as to establish an optimized iterative process and get a feasible collision avoidance path.However,the dynamic partial avoidance mainly considers two kinds of situations: the known dynamic obstacle information and the unknown dynamic obstacle information,and these two cases are dealt with by genetic algorithm and sliding window method respectively.Combined to solve the dynamic obstacle real-time collision avoidance problem.Finally,we simulated and simulated the path planning collision avoidance algorithms under both static and dynamic conditions,and verified the feasibility of the proposed path planning and collision avoidance algorithm.