Autonomous Obstacle Avoidance System Of UAV Based On Millimeter-wave Radar And Monocular Vision

In recent years,small civilian UAV has received more and more attention and application.Unmanned aerial vehicle has been widely used in the fields of express,takeout,power line inspection and so on.The UAV,which has the function of independent obstacle avoidance,has a wide application prospect in social production,scientific research and aerial photography.It can not only make UAVs fly smoothly,but also enable UAV to avoid obstacles effectively when completing various complex flight tasks,and reduce the damage to unmanned aerial vehicles from the obstacles.And the secondary harm brought.Scholars at home and abroad have done a lot of research on the autonomous obstacle avoidance of UAV.The general UAV obstacle avoidance scheme relies on the ready-made 3D elevation map,the multi eye vision sensor or the combination of many other sensors.The obstacle avoidance system based on the existing scheme generally has large volume and high power consumption.The disadvantages of high quality and high price.It has the advantages of small size,low power consumption,light quality and low price,and the independent obstacle avoidance system suitable for installation on small UAVs(such as multi rotor UAVs)remains to be further studied.Millimeter-wave radar has the advantages of small volume,light weight and far ranging distance,but millimeter wave radar has low accuracy when measuring the azimuth angle of obstacles.Monocular vision sensor can measure the azimuth angle of the obstacle in UAV flight environment,and compared with the multiocular vision sensor,monocular vision processing has the advantage of small calculation,but the precision of the monocular vision sensor is very low in measuring the distance of the obstacle.The scheme of millimeter wave radar and monocular vision sensor,which is installed on Small UAV,can obtain high precision obstacle distance and azimuth information at the same time,and realize the advantages of two kinds of sensors.Moreover,the calculation of the scheme is small,and the general embedded microprocessor can be competent.It is installed and used on small UAVs.Based on the autonomous obstacle avoidance scheme of UAV,the main contents of this paper are as follows:1)The research and development status of the autonomous obstacle avoidance system for unmanned aerial vehicles(UAV)at home and abroad are summarized,and an autonomous obstacle avoidance system is proposed,which combines the millimeter wave radar and the monocular vision sensor.2)Introduce the composition of the millimeter wave radar and the basic principle of measuring the distance of the obstacle by using the millimeter wave radar,and study the method of applying the digital image processing theory to the azimuth angle measurement of the obstacle by the monocular vision sensor,and put forward the obstacle of the distance result of the obstacle of the millimeter wave radar and the monocular vision sensor.The method of merging the results of the angle measurement to realize the 3D three-dimensional positioning of the obstacle,and introducing the method of simplifying the complex three-dimensional path planning problem to the two-dimensional path planning problem.3)Based on the DSP hardware platform of TI,millimeter wave radar platform and the Naza UAV platform of DJI,a set of autonomous obstacle avoidance demonstration system which combines millimeter wave radar and monocular vision is developed,and the obstacle avoidance flight test is done using the UAV autonomous obstacle avoidance demonstration system,and the fusion proposed in this paper is preliminarily verified.The feasibility of the scheme in UAV autonomous obstacle avoidance flight is discussed.