AUV Target Search Based On RRT In Unknown 3D Space

In recent years,human research and development of the ocean has become more and more intense,and the problem of target search and moving target interception in the unknown marine environment has always been a hot topic of research.Autonomous Underwater Vehicle(AUV),as a flexible and loadable underwater device,is a technical means to solve the current problem.In this paper,the following research contents are carried out for the target search and interception problem in unknown underwater environment:Firstly,the target search and moving target interception tasks of multiple AUVs in the underwater environment are analyzed,and the solutions for target search and interception are proposed.At the same time,a simulation environment map model,sonar model,AUV motion model and target motion model are established for the elements required to perform tasks in an unknown environment.Secondly,the problem of search path planning in unknown environment is analyzed,and a path planning algorithm based on improved rolling RRT is proposed.The algorithm combines rolling planning,reverse node clipping and sub-target point selection,and uses rolling planning to apply RRT to an unknown environment.In the process of RRT expansion,the randomness of the node clipping algorithm is used to improve the selection of sub-target points.It improves the efficiency of the algorithm,improves the speed of path planning,indirectly reduces the time of the AUV in the search process,and improves the search efficiency.Then,for the target search problem in unknown environment,an AUV target search algorithm based on improved RRT algorithm is proposed.In terms of search,a search decision function including target probability map,uncertainty map,search task cost and regional traversal degree map is established to achieve the purpose of efficient AUV search;in local planning,the improved RRT algorithm is used to carry out Planning,the combination of the two,realizes online real-time search of AUV in three-dimensional space.The attractive principle of artificial potential field is applied to the target search income,and the search of AUV in the dense target environment is realized.Simulation and proof of the neural network algorithm in the dense target environment are carried out,which verifies the performance of the algorithm in the dense target environment.Finally,for the dynamic target interception problem,a position prediction method based on the three-point differential method is proposed,so that the AUV can predict the position of the moving target at the next moment according to the short-term position information after detecting the moving target information,so as to achieve the position in advance.Point and intercept the purpose of success.By analyzing the interception results of moving targets,the searched changes of the AUV search space and the simulation cases of the search results of moving targets and static targets,the effectiveness of the method is proved.