Research And Implementation Of Indoor Mobile Grabbing Robot Based On ROS

Artificial intelligence technology is developing rapidly,and robots are becoming more and more autonomous and intelligent.The integration of robotic arms into mobile robots not only has the function of autonomous navigation but also extends the operation space of robotic arms,which has certain development prospects.Therefore,this paper designs an autonomous mobile grasping robot based on the ROS platform,an open-source operating system for robotics,and studies the key technologies of map construction,positioning and navigation,and recognition and grasping for indoor environments.The overall design of the robot was developed in terms of the functions of each part of the robot.In order to improve the control accuracy,the robot hardware system framework was designed and the sensors were selected,and the integration of the mobile platform and the robot arm platform was completed.In order to facilitate the control of the robot,the software system design was completed based on the ROS platform in the Linux environment.Firstly,the map construction and autonomous navigation algorithm of the robot in indoor environment is studied in depth.Simulations of the Gmapping algorithm based on the Rao-Blackwellized particle filtering principle are performed to confirm the feasibility of the algorithm.By comparing three global path planning algorithms Dijkstra,BFS and A* based on the node search cost,the experimental results show that the A* algorithm has a smaller search cost and shorter time.To address the problems of global paths such as route zigzagging,the paths are optimized by the local path planning algorithm DWA,and the three key evaluation function coefficients of DWA are compared and analyzed to provide a theoretical basis for how to select the coefficients in subsequent robot applications.Then the coordinate transformation relationship between camera,robot arm and target object is established by camera calibration and hand-eye calibration.Compared with the traditional template matching algorithm,the color recognition algorithm has good rotation invariance and scale invariance,so it is applied to the robot target recognition.And the forward and reverse kinematics of the robot arm is solved by the D-H parameter method,while the URDF model of the robot arm is created in the ROS platform,and the motion planning parameters are configured by using the Move It!Setup Assistant tool.Finally,the robotic arm joint space path planning is completed by the cubic polynomial interpolation method,and the validation experiments are completed on the robotic arm.Finally,the map construction,autonomous navigation,and recognition and grasping functions of the robot are verified by building a real indoor scene.The experimental data show that the robot’s positioning and navigation accuracy and recognition and grasping accuracy meet the system requirements,which confirms the high feasibility and reliability of the robot system designed in this paper.