Research On UAV Landing Strategy Based On Lidar

With the rapid development of UAV,its application fields are becoming more and more extensive,so its stable and safe landing problem has also attracted more and more attention.After some drones lose GPS signals or wireless control signals,they cannot land safely because they cannot actively judge the landing environment.In response to this problem,this thesis uses lidar to monitor the landing environment,and finally makes the drone land smoothly.This method is less affected by light,temperature,landform and other environments,and has strong applicability.The research object of this thesis is the distanceless X60 single-rotor UAV.First,the adaptive landing gear is designed according to its structural characteristics,and the Robosense16-line Li DAR and Raspberry Pi are selected to build the UAV landing control platform.Secondly,the data of the landing environment of the UAV is obtained and processed,and the discrete and repeated point clouds are removed using the method based on statistical filtering and the downsampling method of voxel grid respectively.After comparing different point cloud matching algorithms,the normal distribution algorithm is selected for point cloud registration,and the modeling of the environment space is completed.Then,the landing process of the UAV is divided into three stages: hovering,constant speed landing and landing,and the linear height difference method based on the combination of the gradient method and the height method is used to adjust the height threshold to establish a global adaptive model to segment the landing area.Finally,the UAV landing strategy is formulated for static landing environments including planes,slopes,and complex ground,as well as dynamic landing environments with a simulated rocking platform as an example.For a static environment,the height distribution of the landing area is measured by laser,the attitude of the landing gear is adjusted to form a landing gear plane parallel to the landing plane,and the UAV is commanded to land at a constant speed;For the dynamic environment,the wave prediction algorithm based on BP neural network is used to judge the motion state of the rocking platform,and the distance between the UAV and the rocking platform measured by laser is used to calculate the landing time of the UAV and predict the rocking platform attitude when the UAV lands,and adjust the attitude of the adaptive landing gear to form the same attitude as the rocking platform at this time,while commanding the UAV to land at a constant speed.Several landing tests in different environments have been carried out for the above landing strategies.The experimental results show that when the landing strategy is applied to land,the average positioning error meets the landing requirements of the UAV within the preset range.