Research On UAV Autonomous And Precise Landing Technology Based On The Combination Of GPS And Machine Vision

In order to enable the quadrotor UAV to better complete various flight tasks,it is very necessary to realize a high-precision and high-safety autonomous landing system.At present,UAVs rely on satellite navigation systems to initially achieve the purpose of autonomous landing,but problems such as large satellite positioning errors and vulnerability to attacks directly affect the accuracy and safety of autonomous landing of UAVs.However,by adding target detection technology to identify ground landing point markers for assisted landing of drones,compared with the automatic landing system that only uses GPS satellite positioning,it can effectively improve the accuracy of autonomous landing.At the same time,it can use landing markers that carry coded information.Improve the safety of the UAV’s automatic landing system,so as to achieve the purpose of improving the accuracy of the UAV’s autonomous and precise landing.The main work and innovations of this paper are as follows:1.Aiming at the problem of satellite positioning,after analyzing the least squares method and Kalman filtering method in common positioning calculation algorithms,according to the principle that the larger the value of the satellite geometric precision factor,the worse the positioning accuracy is,the least squares method is used to give An improved weighted iterative least square method is presented.In this method,the satellite data is weighted in steps according to the calculated satellite elevation angle value,and the satellite is selected according to the satellite azimuth angle to reduce the weight of satellite data with a larger geometric precision factor when calculating the receiver position.The actual satellite broadcast ephemeris data obtained from an IGS(International GNSS Service,International GNSS Service)data center is compared and analyzed through experiments and it is found that the weighted iterative least square method not only has higher positioning accuracy,but also has stable positioning compared to the least square method.Sex has also improved to a certain extent.2.The ArUco-enhanced mark is proposed on the basis of the ArUco mark,which is used to avoid the detection failure problem caused by the landing point mark on the ground may be blocked by other objects.On the basis of the traditional sliding window algorithm,it is improved according to the distribution data of the landing point mark in the image.Because the drone uses GPS for positioning at the same time,the landing point mark generally appears in the center range of the captured image.On this basis,the detection frame traversal sequence of the original algorithm is changed and the detection frame size is set to a fixed value,which effectively improves The detection speed.Finally,the GPS-based landing system and the combined autonomous fixed-point landing system proposed in this paper are compared and analyzed under software simulation and physical platforms.The experimental results show that the autonomous landing system proposed in this paper is more accurate,and the recognition module based on ArUco mark also helps to improve the safety of the UAV during autonomous landing.