Ground Target Tracking And Soft Landing For Unmanned Multi-rotor Aircraft

With the development of modern technology in recent years,unmanned aerial vehicle(UAV)has attracted more and more attention for its broad application prospects.Various UAV-enterprises appear as bamboo shoots after a spring rain in public.Whether in military or civil field,the intelligence and autonomy are the main directions of development for UAV.In this paper,unmanned multi-rotor aircraft(UAV)is taken as the research object.The ground moving target tracking and soft landing of the unmanned multi-rotor aircraft are studied,based on the electronic technology,image processing technology and aircraft control technology.1.This article takes a quad-rotor unmanned aerial vehicle as the research object and takes the M100 quadrotor aircraft produced by DJI as a development platform.Firstly,aiming at the problem of UAV ground moving target tracking,a multi-work mode solution is proposed.The two working modes are used to meet the task demand under different working conditions,and the overall framework of the system is given in detail from two aspects of hardware and software.The overall framework of the system is designed.Then the modeling and analysis of the quad-rotor UAV is provided,which provides a theoretical basis for the content of the subsequent chapters.2.For the problem of large-scale target tracking of quadrotor UAV,this paper proposes a scheme,which track ground target by electronic tags and designs the structure of electronic tags.Then we further study the control problem of the M100 unmanned aerial vehicle under this working mode,and test the proposed scheme under the hardware-in-the-loop simulation environment by DJI and outdoor practical experiment.The experimental results verify the effectiveness of the program.3.Aiming at the problem of visual tracking of quadrotor UAV,this paper designs a visual tracking system based on the classic Cam-shift algorithm,which includes the target image coordinate transformation,the hardware and software implementation of the system.Then,aiming at the deficiencies of Cam-shift algorithm,this paper further studies the target-tracking algorithm(KCF algorithm)based on nuclear correlation filtering.The actual tracking experimental results prove the effectiveness of the visual tracking system of UAV designed in this paper.4.For the soft landing problem of quadrotor UAV,this paper first solves the problem of inaccurate positional information collected by the vision system.A method based on curve fitting filter is used to process the original data.On the one hand,this paper studies the optimal landing trajectory problem based on the cooperation goal,and deduces the optimal control rate with the attitude of the landing attitude as the constraint condition.The validity of the proposed method is verified by MATLAB simulation and outdoor practical experiment.