Design Of Robot Autonomous Sorting System For Intelligent Logistics

In recent years,with the rapid development of the logistics industry,there are more and more kinds of goods in the warehouse,which leads to the complex sorting task and requires a lot of labor.The ap plication of vision based manipulator technology in the logistics sorting task can solve the above problems.The robot arm based on vision is easily affected by the external environment,so how to improve the intelligent level of the robot arm in the complex environment(light changes,obstacles)has become a hot issue.Based on this,this paper uses ROS as the development platform and ABB IRB120 manipulator as the operation platform to design a set of logistics industry oriented manipulator autonomous sorting system,which can meet the real-time and accuracy requirements of sorting tasks in complex scenes.The main research work and achievements are as follows:(1)Aiming at the problem of robot's autonomous sorting of unknown objects,the paper designs a vision based industrial robot's autonomous sorting system.The m ain operation mechanism of the scheme is as follows:firstly,the camera captures an image and uses the object detection algorithm based on deep learning to recognize and detect all the objects on the robot operation platform;Secondly,according to the recognition and detection results,the target is selected by human-computer interaction;Then,only the target chosen by the user is detected by applying the target detection algorithm to get the present image features of the captured target;Finally,the 3D coordinates of the object in the robot base frame arc obtained by image registration and coordinate transformation;Based on this,the 3D coordinates of the object to be grasped are sent to the manipulator controller,and finally the manipulator is controlled to complete the sorting task on the console according to the trajectory planning algorithm.Through the corresponding experimental research,it is proved that the system can meet the requirements of millimeter level.(2)For the calibration issue of robot vision system made up of Intel RealSense D435 depth camera and ABB IRB120 manipulator,at first,the paper calibrates the interior parameter of the camera by adopting the camera_calibration function package according to Zhang Zhengyou algorithm and obtains the internal parameter matrix and distortion parameters under ROS;and the internal parameter matrix and distortion parameters are obtained;Then the parameters(type of equipment,parameter of calibration plate,IP)of the equipment involved in the visual calibration easy_handeye function package are configured,the robot vision system composed of camera and manipulator is calibrated(also known as hand_eye calibration),and finally the conversion relationship between camera coordinate system and manipulator base coordinate system is obtained.(3)Aiming at the problem of three-dimensional position estimation of multiple objects(no overlap)randomly placed on the operating platform,this paper first uses Yolo v3 algorithm to obtain the image features of the target to be captured,then carries out image registration on the color map and depth map collected by the camera,and finally obtains the 3D coordinate information of the target in the base coordinate system of the manipulator combined with the coordinate conversion technology.(4)In order to solve the motion control simulation problem of the manipulator,the paper first creates the manipulator model through URDF file in ROS,and configures the communication files of RobotStudio(Server)and ROS(Client)respectively;Secondly,the communication between them is established through the Socket interface based on TCP/IP;Then linear interpolation algorithm and circular interpolation algorithm are used to plan the trajectory of the manipulator in Cartesian space;Finally,the robot arm of RobotStudio is simulated and controlled by MoveIt!in ROS.(5)In practical application,the communication between ABB manipulator and PC depends on RobotStudio software.The autonomous sorting system designed in this paper abandons the communication mode of RobotStudio software.Firstly,it configures the communication files related to ROS communication and the files defined for trajectory and joint point data to the entity controller,and then establishes communication with the entity robot arm through ROS,Finally,the experimental platform of the whole autonomous sorting system is built,and 12 groups of experiments are set up to quantitatively evaluate the hand_eye calibration results and the performance of the algorithm.In the meantime,6 groups of physical experiments are set up to conduct the automatic sorting experiments on a large number of commodities in the storage and express boxes in the simulated conveyor belt environment.According to the analysis of the experimental results,the accuracy and grasping success rate of the manipulator autonomous sorting system meet the requirements of the designed system,which proves the feasibility of the system.