Design Of Mobile Manipulator’s ROS Simulation System Based On Vision

Traditional industrial robotic arms have been widely used in the production field,but limited by the fixed base,the working space of the robot is limited.In this paper,a mobile visual manipulator simulation platform is constructed,which realizes the function of the manipulator to grasp the target object by vision and the mobile chassis to navigate autonomously.The processing method of the trajectory of the object grasped by the robotic arm is preliminarily explored,and the processing flow of the image information obtained by the camera and the path planning of the mobile chassis are discussed.It provides a simulation platform for future research.With the help of the unique operating mechanism and technical characteristics of the ROS robot operating system,the code reuse in the robot development process can be improved,and the coupling between various functions can be reduced.ROS also provides many tools and plug-ins,which can be very convenient and concise to solve the problems of trajectory planning of robotic arms and autonomous navigation of mobile chassis.For the robotic arm to perceive the spatial coordinates of the target object,use the Kinect depth camera to obtain image information.The template matching method is used to determine the position of the target object in space,and then the position of the object relative to the robotic arm is determined through coordinate transformation.With the help of the Move It! tool provided by ROS,the trajectory of the robot arm is planned,and the robot arm is controlled to complete the grasping of the target object.The mobile chassis builds the environment map using the Gmapping method,and encapsulates the corresponding function package in ROS.When calling the function package,modify the corresponding parameters to complete the exploration of the surrounding environment.For path planning,the principle and execution steps of Dijkstra and A~* algorithms in move＿base are analyzed.