Research And Design Of Mobile Robot Autonomous Work System

In recent years,mobile robots equipped with multi-degree-of-freedom manipulators have been widely used in areas such as anti-terrorist detonation,disaster relief,space exploration,national defense,military and other fields to assist humans in performing dangerous operations.However,the manipulation of multi-degree-of-freedom robotic robots is relatively complicated.Therefore,studying autonomous operation system and improving the degree of autonomy and intelligence of robots are the hotspot in the research field of robot control systems.In view of the above hotspot,a set of mobile robot autonomous operation system was designed,which makes the robot have the ability of autonomous grasping of target objects,improves the success rate and speed of grasping,and makes up for the deficiency of manual operation.The main contents of this paper are as follows:(1)Aiming at the need of the six-degree-of-freedom explosion-proof robot to grasp objects accurately and quickly,based on the stereo depth camera as the visual sensor,the problems of environment perception,grasping pose detection and recognition,visual SLAM and so on are studied and experimented.assisted grasping,autonomous grasping and other functions are realized.(2)The overall design of the mobile robot autonomous operation system is designed from the hardware and software level,and then the calibration of the two stereo cameras is completed on the theoretical basis of the small hole imaging model to obtain a three-dimensional point cloud image and implement the auxiliary grasp function.The geometric shape of the robot arm is modeled,and the related geometric information is extracted to filter sampling points so as to get a reasonable set of grabbing poses.Different weights are given to the grasp poses by neighborhood search,and finally the grabbing pose with the highest weight is selected as the best grasp pose,realize the function of autonomous grasping and simulation experiments are conducted.(3)Based on the ROS architecture,design the autonomous operation system software solution,write the modular autonomous operation core function nodes.Use the QT framework to write the autonomous operation system's human-computer interaction interface.Select objects with typical shape characteristics,and design the manual grasp,assisted grasp,autonomous grasp multiple sets of experiments,and visual SLAM test experiments.The results show that the autonomous operating system can control the robot to quickly and accurately grab the target object,which can satisfy the application needs.