Research On UGV Path Planning Technology Based On Top Views Taken By Cooperative UAV

Based on the help of environmental-aware and path planning system,unmanned vehicle can replace people to complete heavy or dangerous tasks such as environmental exploration,cargo handling and safety detection in special environment.At present,the environment-aware system used by unmanned vehicles has problems such as obstacle occlusion and limited detectable area,and the path planning algorithm used has problems such as the serious fit between path and obstacle.Aiming at these problems,this paper studies the path planning algorithm of unmanned vehicles based on the top view taken by cooperative unmanned aerial vehicles.1.The image mosaic and image segmentation algorithms of top view taken by cooperative unmanned aerial vehicles are studied.Firstly,the unmanned aerial vehicles is used to take the top view of the environment for the unmanned vehicles,then the image mosaic algorithm is used to generate the panoramic environment map,and then the image segmentation algorithm is used to extract the feasible region and infeasible region in the panoramic environment map.Finally,the obstacles are loaded for the environment map according to the infeasible region and the preset forbidden area.It solves the obstacle occlusion problem in the environmental-aware system and the problem of crossing the forbidden area in the path planning.2.The fusion algorithm of image thinning and global path planning for unmanned vehicle is studied,and the path optimization algorithm is improved.Firstly,the grid method is used to preliminarily establish the environment map,and then the image thinning algorithm is used to refine the feasible region in the environment map.Finally,the global path is planned for the unmanned vehicle according to the refined feasible region.It solves the problem of too close between the global path and obstacles and too many path turning points.3.The real-time local path planning algorithm for unmanned vehicles is studied,the scheduling methods of global path planning algorithm and local path planning algorithm are designed,and the artificial potential field algorithm is improved.Firstly,the planned global path is segmented according to the turning point,and then the improved artificial potential field algorithm is used to plan the local path for each segment to guide the unmanned vehicle to move to the target point.The problem that the artificial potential field method will fall into local minimum and U-shaped obstacles is solved.4.At the system of Ubuntu 18.04,the system simulation experiment platform is built with Qt5.13 application development framework.The corresponding experiments are designed for the proposed and improved algorithms in this paper,which are tested and verified in the system simulation experimental platform,and the experimental results and data are given.The effectiveness and practicability of the proposed and improved algorithm are verified in the real scene.The experimental results show that:(1)the environment-aware system based on the top view taken by cooperative unmanned aerial vehicles proposed in this paper can perceive other obstacles blocked by obstacles;(2)The global path planning algorithm based on refined grid map proposed in this paper can plan a safe path away from obstacles;(3)The improvement of the artificial potential field method in this paper can make the algorithm escape from local oscillation points and U-shaped obstacles.