Algorithm And Implementation Of High Precision SLAM For Laser Mobile Robots

Simultaneous localization and map building(SLAM)are two basic and important problems in the relative navigation of mobile robots.The positioning of mobile robot depends on the environment map with sufficient accuracy,and the map is created based on the exact positioning of robot.In addition,the real-time localization and map creation(SLAM)ability of the robot is a necessary ability for the mobile robot to move in the unknown environment,and it is also an important basis for the complete autonomous operation of the robot.The robot starts from the unknown position in the environment,estimates its own pose during the movement,and builds an incremental map of the surrounding environment through sensor measurement,and updates the robot pose with the map.The positioning of mobile robot requires the construction of environment map.The construction of the environment map also depends on the pose of the robot,and the positioning of the robot is complementary to the environment modeling.SLAM concept is the key technology to solve this problem.However,most of the current research is still in the laboratory research stage,and the successful implementation of real,large-scale environment is relatively rare.In addition,the robot application environment in the industrial field is more complex,requiring higher robustness and positioning accuracy.However,there is a large error between the two-dimensional map of the main construction environment and the real environment,which will directly lead to the application of industrial robots in the real scene is greatly limited.In such a complex industrial environment,there are many uncertainties that hinder our SLAM problem.The problem of "false landmarks"is one of the most serious obstacles:first,the probability of a mismatch between a real landmark and a false landmark may increase,which could lead to a difference in SLAM algorithms based on EKF.Second,these unreliable"false landmarks"are always unstable,leading to more uncertainty and even divergence.Third,as more and more"false landmarks"bring higher dimensions of state vectors and covariance matrices,the computational complexity of EKF-based algorithms may increase dramatically.Therefore,this study aims to establish the environmental model of complex industrial scenes with 2D radar,and propose a dual-map-based anti-jamming EKF SLAM framework in the practical environment.The framework adopts strict landmark selection method and feature filtering mechanism based on double maps to eliminate"false landmarks",effectively reduce "noise" and improve the positioning accuracy of the construction map.Meanwhile,the landmark processing and real-time observable area are optimized to further ensure the correctness of the landmark.Finally,the experimental errors before and after the improvement are analyzed and summarized.And the validity and practicability of the improved algorithm are verified.