Research On Obstacle Detection And Path Planning Technology Of Quadrotor UAV

Multirotor unmanned aerial vehicle(UAV)has been widely used in many fields of daily life,such as survey,photography,plant protection,logistics,etc.,which bring great convenience to people's life.More and more research institutes and enterprises begin to research the autonomous flight technology of UAV,and obstacle avoidance ability is an indispensable part of it.The development of UAV obstacle avoidance technology,on the one hand,can improve the reliability and safety of UAV flight,on the other hand,it can reduce the risk of personal injury and property loss caused by UAV.Therefore,the research of UAV obstacle avoidance technology has important theoretical significance and application value.Considering that most of the existing obstacle avoidance technologies have problems such as poor performance of obstacle detection and slow response of obstacle avoidance,this paper discusses the design and implementation of obstacle avoidance technology for quadrotor UAV,and the main contents are as follows:Firstly,from the system point of view,the system structure and the function of each component unit of the quadrotor UAV are analyzed.Through the force analysis of different flight attitudes of UAV,the control principle of UAV in flight is described,and the actual flight experiment is completed.Secondly,an obstacle detection technology based on image information and lidar information fusion is designed.The training of Yolov3 is completed by using the local data set,and the obstacle area information is extracted by using the trained model.At the same time,the scanning lidar is added to obtain the obstacle distance information.The transformation relations among pixel coordinate system,image coordinate system and camera coordinate system are established to realize the 3D mapping of obstacles from image to space.Furthermore,on the basis of known global environmental information,a global path planning strategy based on the fusion of Dijkstra and improved particle swarm optimization algorithm is proposed.The Dijkstra algorithm is used to search the path,and the particle swarm optimization algorithm is combined to improve the path search efficiency.At the same time,on the basis of traditional particle swarm optimization algorithm,multi-dynamic population strategy and nonlinear attenuation inertia weight strategy are introduced to improve the algorithm.MATLAB simulation results show that the proposed strategy can generate a global optimal path quickly and efficiently.Finally,under the condition that the global environment information is unknown,a localpath planning strategy based on the obstacle area information is proposed.Using the obstacle detection technology designed in this paper,the area information of the obstacle is obtained,and the corresponding obstacle avoidance action is made according to the relative position of the UAV danger window and the obstacle boundary box.The simulation results of ROS show that the proposed strategy can help the UAV to bypass different obstacles in real time and ensure the flight safety of the UAV.