System Design And Research Of Mobile Robot Hand-Eye Coordination Arm

Intelligent robot almost walks in all aspects of human's life,and ubiquitous intelligent robot is changing people's lives.Due to its multiple functions and widely usage,mobile manipulator is widely used in industry,service,special field,military and aerospace fields.With the widening of the application field of the mobile manipulator,the operating environment it faces is increasingly complex,Therefore,the research on autonomous grasping system of mobile manipulator in unstructured environment is becoming more and more popular.The main contents of this paper are as follows:(1)The research status of autonomous grasping system of mobile robot arm was analyzed,and some key technologies of the system were summarized.A five degree of freedom manipulator and the overall hardware system architecture of the mobile manipulator were designed.Some hardware devices were selected and tested such as joint motor,encoder,remote control handle,IPC,bottom controller,camera,touch screen and power supply module.Finally,the control system and software system architecture of the mobile manipulator were built,and the connection between ROS and other devices was also configured.(2)Initially,the forward kinematics was analyzed base on D-H method.Then the motion space of the manipulator and the analytical solution of the inverse kinematics were obtained.Finally,both the kinematics performance and dynamics performance of the manipulator were also simulated and analyzed.(3)Firstly,autonomous grasping software system architecture was set up.Then the server parameter of the manipulator was built by MoveIt!,and the algorithm of joint space trajectory based on random sampling was also studied and applied.Besides,the closed solution of the manipulator was also configurated.Finally,the joint controller was designed to control the actual manipulator.An experiment was conducted to verify the reliability of the closed solution.(4)Initially,the color feature and semi-global disparity were used to detect and calculate by using the binocular camera module.The transform relationship between the camera coordinate system and robot coordinate system was calibrated by using the least squares method.An online recognition and grasping experiment of objects was conducted,which verified the feasibility and accuracy of the target detection and automatic grasping.(5)A task scheduling system which integrates autonomous navigation,object recognition and autonomous grasp was put forward and a task behavior tree was set up under the ROS system.Action type and task scheduling interface of the visual function and grasping function were also designed.Actual sences were built to conduct the physical testing of task scheduling system,which verified the reliability of the autonomous grasping system further.