Research On Indoor SLAM Technology Based On Cartographer And RBPF

As one of the branch of artificial intelligence,intelligent robot technology integrates traditional theories such as mathematics,automation,machine mechanics,etc.Its achievement will be along with the frontier disciplines of artificial intelligence such as vision and deep learning.The realization of the "autonomy" of the mobile robot movement is the shape of things to come,and mobile robot simultaneous localization and mapping(SLAM)is the premise of the robot's "autonomy".However,the current SLAM technology faces a series of problems such as low accuracy of the constructed map,poor real-time performance,and confusion of application scenarios of various algorithms,resulting in SLAM not realizing productization.In view of the existing problems,the accuracy of SLAM algorithm and the application scene of SLAM algorithm are studied in this paper.Firstly,the principles of RBPF and Cartographer algorithm are studied in this paper,and the real-time performance of RBPF algorithm is improved accordingly.The accuracy of the maps constructed by the RBPF algorithm,the improved RBPF algorithm and the Cartographer algorithm is simulated and analyzed.It is concluded that the real-time performance and accuracy of the map created by the improved RBPF algorithm is better than that of the standard RBPF algorithm.The simulation results also prove that the improved RBPF algorithm is superior to the current most popular Cartographer algorithm in the speed of building maps,and the difference of the accuracy in the small scene between the improved RBPF algorithm and cartographer algorithm is very small.In order to further verify the positioning accuracy of the improved RBPF algorithm,the experimental of the positioning accuracy of the improved RBPF algorithm and the Cartographer algorithm is designed in the real enviroment.Secondly,for the mobile robot used in the experiment,the differential motion model of the robot and two methods for calculating trajectory are stuided: dynamic window method and two-wheel differential method.Experiments show that when the distance traveled by the mobile robot used in the experiment is 20 meters and dynamic window method and two-wheel differential method are used to calculate the robot trajectory,the cumulative error of the dynamic window method is slightly smaller.the dynamic window method is used to calculate the robot's mileage information.Finally,aiming at positioning accuracy of the Cartographer algorithm and the improved RBPF algorithm,a mobile robot experimental platform based on ROS and SLAM algorithm is set up.The functions of autonomous movement,map building and positioning,the real-time monitoring of mobile robot are realized.In the experiments,the deviations of the map positioning using the two algorithms in short-path,long-path and complex environments are collected respectively.The results show that the improved RBPF algorithm has the same accuracy as the Cartographer algorithm in both short-path and long-path environments.The Cartographer algorithm has an advantage in a complex environment.Combined with the advantage of the improved RBPF algorithm in the speed of map building,an improved RBPF algorithm is proposed for building maps,the Cartographer algorithm is recommended for positioning in a complex environment,and the improved RBPF algorithm and the Cartographer algorithmin all can be used in a small and single-structured environment.