Research On The Control System Of Aerial Manipulation Based On Visual Servo

In recent years,rotorcraft UAV technology has been rapidly developed and has been widely used in various industries.However,its application is mostly in areas that need to avoid contacting with the outside world and have no interaction,so it does not fully play the function of UAV.Aerial manipulator system is composed of the UAV and the series or parallel manipulator with operability that is loaded on the abdomen of the flight platform,which endow the system with the ability to interact with the outside world and achieve aerial manipulation.The system combines the mobility of the UAV in three dimensional space with the flexible operation of the manipulation,extending the work range of the traditional rotorcraft UAV.This thesis studies the aerial manipulator system with air autonomous operation ability.The main work is to design an appropriate controller for the aerial manipulator system.The improved image visual servo control method and multi-level task priority control method are adopted to realize the autonomous positioning of the end of the arm above the captured target.The details are as follows:(1)Joint modeling of aerial manipulator.In this thesis,the dynamical model of quadrotor is introduced first,then the structure of the manipulator is designed reasonably,and the kinematics of the manipulator system is modeled.Aiming at the center of gravity offset of the system caused by the movement of the manipulator arm,the center of gravity offset vector is introduced,and the joint modeling of the aerial manipulator system is realized based on newton-euler equation,in order to further realize the stability control of the system.(2)Improved perspective projection image features.This thesis introduces in detail the perspective projection image features and spherical projection image features involved in the classical image visual servo.In order to solve the problem of nonlinearity and coupling between image space and cartesian space in the motion of aerial manipulator system,an improved perspective projection image feature,namely virtual image feature,was proposed in this thesis.The dynamic equation of image feature is derived by combining image moment.(3)Adaptive visual servo controller based on virtual image features.In this thesis,a spherical projection based visual servo controller and a virtual image based visual servo controller are designed respectively by backstepping and adaptive backstepping.In the design of visual servo controller based on virtual image,the adaptive parameter adjustment mechanism is introduced,aiming at the uncertainty of image information and the uncertainty of system dynamics model,the adaptive adjustment can improve the accuracy and robustness of the system.(4)Multiple task priority control.Then,it realizes the coordination control of the UAV and the manipulator arm.For the design of the manipulator controller,this thesis proposes a priority control method based on the control of position,direction and center of gravity.Ensure that the end of the manipulator is positioned in the specified position while ensuring the stability and safety of the system.In order to verify the control method proposed in this thesis,a simulation model of the aerial manipulator system and controller were built.Results show that these visual servo control strategies have good performance,through coordinated control UAV and manipulator arm,the end of the arm can be automatically positioned to the captured target.At the same time,the simulation results show that the proposed visual servo control method based on virtual image features has better performance in space and image plane than the visual servo control method based on spherical projection,and the convergence time is also improved.At last,the aerial grab experiment in the simple outdoor environment was carried out by using the aerial manipulator system independently designed by the research group.