| The autonomous landing technology of rotor UAV for moving platform integrates target detection,autonomous decision-making and flight control technology,which has been widely used in military,civil and other fields,and has important research value.Through the preliminary study,it is found that there are three main problems in this technology: first,the location of the moving landing point is uncertain,and the angle of view of the airborne camera is constantly changing during the landing process of the rotorcraft,which makes it difficult to detect the landing point;Second,the traditional detection algorithm is inefficient and can not run in real time on the airborne computing platform;Thirdly,in the process of autonomous landing,the flight control deviation of the rotor UAV is large,which makes the UAV unable to land on the moving platform accurately.In view of the above problems,the research work of this thesis is mainly divided into three aspects.Based on Apriltag algorithm,a multi-directional marker is designed to guide the rotor UAV to land accurately.The marker is composed of the middle nested April Tag code and the surrounding distributed April Tag code,which provides more effective information,so that the rotor UAV can accurately locate the landing point at different heights and directions,and solves the problem of missing moving landing point detection.On this basis,aiming at the low computational efficiency of Apriltag algorithm,an acceleration strategy is proposed to reduce the traversal area.DSST tracking algorithm is used to track the multi-directional markers,and the tracking results are used as the detection area of Apriltag algorithm to reduce traversal and complete acceleration.Experiments are carried out on the airborne computing equipment,and the experimental results show that the proposed method solves the problems of marker loss and low computational efficiency in the detection process.The autonomous landing control strategy of rotor UAV Based on PID control algorithm and ADRC control algorithm is designed.Aiming at the poor dynamic tracking performance of rotorcraft,PID control algorithm is adopted in horizontal direction to improve the response speed of rotorcraft tracking moving landing point;In the vertical direction,ADRC control algorithm is used to solve the overshoot problem in the height descent control and ensure the safety of the landing experiment.At the same time,the extended state observer ESO of ADRC control algorithm is used to continuously observe the disturbance and compensate,so as to improve the antiinterference performance of the system.The control strategy improves the overall dynamic tracking performance of rotor UAV.The feasibility of the control strategy is verified by the design and simulation experiments,which can be well applied to the autonomous landing scene of rotor UAV.Based on the M100 rotorcraft experimental platform,the UAV autonomous precision landing system is built,and the above methods are verified by experiments.In the real scene,the experiments of autonomous fixed-point landing of rotorcraft and moving platform landing are carried out respectively.The experimental results show that the system can identify and locate the moving landing points,control the rotorcraft quickly and accurately,and finally realize the autonomous and accurate landing of rotorcraft. |