Research On Algorithms Of Localization And Mapping Of Mobile Robots In Long-narrow Tunnels

With the improvement of the intelligence and autonomy of mobile robots,the use of mobile robots to perform tasks in underground scenes such as coal mine tunnels and subway tunnels has the advantages of high efficiency and no fear of danger.However,underground scenes have complex working conditions,poor road conditions,no GPS,and poor communication,which pose huge challenges to mobile robot positioning and mapping,especially in long and narrow underground tunnels,there are degraded scenes with few features.Common ground positioning methods cannot be directly applied.In order to solve the above problems,this paper adopts the fusion scheme of lidar,ultrawideband,and inertial measurement unit,and proposes a method for positioning and mapping of mobile robots for underground narrow and long tunnels.The main work is as follows:(1)Establishment of mobile robot positioning and mapping platform system.First build a mobile robot software and hardware platform,select the sensors needed for positioning and mapping based on scene features,and then combine the description of the robot’s motion state to further establish the observation model of the mounted sensors,study its applicability,and provide a platform for subsequent algorithm research and experimental verification basis.(2)Research on the laser SLAM algorithm for the underground narrow and long tunnel environment.First introduce and analyze the robot state estimation method,and then compare and analyze the three most widely used laser SLAM algorithms in principle.Furthermore,the adaptability and reliability of the three algorithms are studied through indoor long corridors and long tunnel experiments.Through analysis,it is found that the existing algorithms have limitations on long and narrow tunnels,and cannot well deal with the cumulative error caused by sensor degradation caused by long and narrow tunnels and bumpy roads.(3)Aiming at the problems of existing positioning algorithms in the context of this article,a method using UWB positioning system assistance is proposed to improve the accuracy of mapping.Based on UWB wireless positioning technology,the EKF-based method is used to fuse IMU observation information to construct a UWB positioning system.At the same time,experiments are designed to verify its accuracy and applicability in the underground narrow and long tunnel environment.The experimental results show that the average positioning accuracy of the proposed positioning system in a long narrow tunnel is 22.3cm(4)Build a laser-ultra-wideband fusion SLAM system.First,use a graph-based optimization framework to add the position constraints provided by the UWB positioning system to the pose graph optimization constraints to provide a reliable initial estimate for scan matching.Then,through point cloud de-distortion,radius filtering,and voxel grid filtering,etc.The original point cloud preprocessing method combines the optimized pose for accurate pose estimation and mapping.Finally,it is proved that the proposed scheme is effective through the experiment of real underground narrow and long tunnel.This paper has 54 figures,17 tables and 74 references.