Research On SLAM Technologies Of Mobile Robot System Based On ROS

In recent years,with the breakthroughs in artificial intelligence,computer technology and sensor technology,the research and development of autonomous mobile robots has made much progress.The mobile robot is no longer limited to the laboratory environment,some service robots that can navigate and avoid obstacles autonomously indoors have been appearing,and even appearing some intelligent air robots which can realize the fast target tracking and the obstacle avoidance.All these are inseparable from the robot's perception of the environment and its self-location.These two technologies are known as simultaneous localization and map-ping(SLAM),which are the key technologies of autonomous mobile robot system.It is also a research hotspot in the field of mobile robots in recent years.Aiming at indoor autonomous mobile robot,this paper develops a ROS based mobile robot system and SLAM technology research.First,the research and experiment of SLAM system are carried out.An improved KinectFusion algorithm loosely coupled with inertial measurement unit is proposed,unfolds a stronger robustness without obvious real-time perfor-mance loss than original algorithm.This lays a foundation for the construction of the mobile robot SLAM system.Then a more complete RGB-D SLAM system has been built in the ROS system,including visual odometer,global optimization and loop detection,which can prelimi-narily meet the environment perception and self positioning requirements of the mobile robot,and the dense map can be used for the navigation and obstacle avoidance.The construction of mobile robot system takes advantage of the characteristics of ROS system,and realizes networking,distributed and modularization.The system includes a un-derlying controller and the upper software system.It can realize remote real-time monitoring and control,data acquisition and transmission,and local path planning and obstacle avoidance.Through the experiment,the SLAM system and the robot system have been tested.In the unknown environment,the real-time simultaneous localization and mapping,and the local navigation and obstacle avoidance of the mobile robot are successfully realized.The feasibility of the SLAM system and the mobile robot system based on the ROS design is proved.