Tracking Control And Platform Construction Of The Networked Manipulator

In recent years,with the rapid development of science and technology,computer vision and robotic arm technology play vital roles in more and more industries.The project of target location is not only the important research content in computer vision field,but also the very challenging research in other areas.In the visual navigation,public scene monitoring,intelligent display visual guidance,intelligent transportation,military museum,and many other fields,target location has very broad prospects.Recently,the theory and method of target location become mature after constantly improving,making the researches based on visual positioning have the solid theoretical support.The research of visual location is of great significance.Based on the actual requirements of target recognition and real-time tracking,this dissertation studies the networked manipulator tracking control and the construction of its platform.The goal of the platform construction is to track the target object on the plane in real time and predict its final position,and control the manipulator to reach the designated position.Firstly,this paper introduces the modeling of the manipulator and its forward and inverse kinematics,and then studies the camera model,making a detailed introduction of Zhang's plane template calibration method.Then this paper introduces the image distortion problem.Secondly,this paper combines shape detection with color detection to identify the target object.The method of superposition between camera coordinate and 3d world coordinate is put forward,then deriving the coordinate transformation formula of graph coordinate and 3d coordinate,which can be used to obtain the 3d coordinate of object accurately.The last position of the object is predicted by polynomial fitting with RLS method.Finally,combining the software development and hardware equipment,a manipulator platform based on monocular vision is designed,and the overall experimental process.The experiment achieved the expected results,which shows that the method is feasible and the solution of inverse kinematics of the manipulator is correct.The difficulties in this research include the processing of network delay,camera calibration,single visual distance,target recognition and detection,and the motion planning of the manipulator.Moreover,the innovation of this paper lies in the proposed visual positioning method,predictive tracking algorithm and the inverse kinematics solution algorithm of the manipulator.The visual positioning based manipulator control system designed in this paper has the advantages of simple operation,strong stability,convenient installation,low cost of platform construction and easy maintenance.