Research And Application Of Robot Mapping And Navigation Technology Based On ROS

With the rapid development of modern social science and technology,artificial intelligence technology is at the forefront of scientific development,Mobile robots are emerging in daily life and continue to develop in a more flexible,error free and intelligent direction,and robot related research has received widespread attention in the development of science and technology at home and abroad.Among them,the Simultaneous Localization and Mapping and Navigation functions of mobile robots are widely used in power inspections and service guidance.SLAM(Simultaneous Localization and Mapping,SLAM)and path planning based on ROS of robot is studied in this paper,and use the ROS platform is used for related research.Firstly,experiments and research concerning localization and mapping are carried out to realize accurate robot self-localization and environmental mapping in unknown and unfamiliar environment.In this paper,the traditional RBPF-SLAM method is studied,which is based on Bayesian filtering and sequential importance sampling to separate localization and mapping.Aiming at the problem of particle degradation in this method,an improved RBPF-SLAM method is proposed.The odometer data is selected as the observation model,and the effective particle threshold and effective particle number,reduce the number of resampling,thus improving the accuracy of robot pose estimation and mapping.Secondly,the path planning of mobile robot is studied to achieve the safe and collision-free arrival of the robot from the initial position,which is actually a dynamic path planning problem of mobile robot.algorithm and ant colony algorithm in global path planning,artificial potential field method and dynamic window method in local path planning are studied.MATLAB is used to simulate the algorithm.Finally,algorithm and dynamic window method are selected to complete the path planning of mobile robot.Finally,obstacle avoidance navigation is implemented.The robot starts to move from the starting point,and makes real-time positioning and mapping in the process of walking according to the simulated real scene.Then,in the built map,the robot is assigned a target point,and the rviz is used for observation.Experiments show that the robot can reach the target point accurately.Then set obstacles in the built map,and the robot can avoid obstacles,reach the target point,complete the robot obstacle avoidance experiment,realizing the navigation process.The algorithm used in this paper realizes the localization,mapping and obstacle avoidance navigation of mobile robot,and has been applied to AGV robots.