Research On Control Of Shipborne UAV Recovery System Based On Mechanical Arm

Unmanned aerial vehicle(UAV)and unmanned surface vessel(USV)are two important types of unmanned systems,which have wide applications in military,civil and scientific fields.Currently,the final landing stage of UAV mainly adopts optical flow sensors and image recognition to assist in positioning and landing.The UAV control algorithm relies on a large amount of target databases,which requires high computational load and is difficult to achieve real-time processing.Therefore,developing a system that can automatically recovery UAV has broad prospects and practical value.In this paper,an automatic recovery UAV system based on a six-axis mechanical arm is designed,and the six-axis mechanical arm is used as the research object to establish the compensation control algorithm and the visual positioning algorithm.The main research contents of this paper are as follows:Establishing the linkage coordinate system of the manipulator,using the standard DH parameter method to establish the kinematics model of the manipulator,deriving the formulas of forward kinematics and inverse kinematics of the manipulator,and analyzing the working space of the manipulator to provide a theoretical basis for the compensation control of the manipulator.Cartesian space trajectory planning is carried out to construct the shortest position and attitude smooth change trajectory in the task space.Research the modeling and calibration methods of the manipulator vision system,including camera imaging model,camera calibration,hand-eye calibration mode,and obtain the precise transformation matrix relationship of the manipulator vision system.Through visual image processing algorithms such as image preprocessing,perspective transformation,image threshold segmentation,and marker detection,completing the detection and recognition of Ar Uco code markers,combing with Pn P algorithm to calculate the pose of the marker.Using AQWA software to simulate the motion of USV,and based on the simulation result data,the long-short-term memory neural network is used to predict the motion trend of the USV.Establish the motion compensation control algorithm model of the manipulator,capture and recover the UAV control strategy,and build a physical model in MATLAB/Simulink for simulation verification.Designing the experimental system scheme of the manipulator to recover the UAV,completing the manipulator control program of the ROS host computer system,and writing the program to control the manipulator through the Python development interface of the Move It! and Actionlib action planning library.Finally,the motion of the USV is simulated by the six-degree-of-freedom motion platform,and the motion compensation control experiment of the manipulator and the recovery of the UAV is carried out to verify the effectiveness of the control algorithm.