Research On Indoor Navigation And Object Grabbing Strategy Of Crawler Mobile Robot

With the development of robot industry,robots come to industry and people's daily lives,replacing humans to complete repetitive and dangerous work.Recently,artificial intelligence and robot technology have merged with each other,which has made tremendous changes of robot perception,calculation and decision-making.In this paper,a differential steering mobile robot is used as carrier to study mapping and navigation,and manipulator grab is realized by image processing.Firstly,introduce hardware and software platform.Laser radar,Attitude sensor and depth camera are main sensors.Beckhoff TwinCAT3 and ROS systems are used as software modules.The speed relationship between robot and wheel is analyzed,and the linkage performance of the wheel-swing coupling motion is studied to optimize the performance.Simulate motion of 5-DOF manipulator in MATLAB's Robotics Toolbox,derive equations for positive and inverse kinematics,and implement the inverse solution through the C++.Calculate all the joints and choose an optimal solution.Secondly,introduce SLAM technology and analyze the principle of Cartographer.Control the robot to build a map of the environment.To realize autonomous navigation,the 2D map is first transformed into a cost map,and the optimal path is planned through navigation stack.According to amcl system,estimates the current robot pose,calculates the speed,and delivers it to robot.Finally,manipulator grabbing is achieved by image processing.Firstly,imaging principle and camera calibration process are introduced.Then the color and depth image are read by depth camera,and then filtered.In the target frame of the artificial painting,the foreground is extracted by GrabCut.Then the foreground is binarized,and the centroid of the target object is calculated.Finally,robot arm is controlled according to target position and kinematic inverse solution algorithm.According to hardware and software platform and above-mentioned algorithm,autonomous navigation and manipulator grabbing experiments are carried out to verify the feasibility and rationality of the system.