Aiming at the problem that the degradation of long pipe gallery features leads to difficulties in mapping

navigation and localization for inspection robots

a map-building and navigation and localization method for inspection robots under weak features is proposed.Firstly

a point cloud map is obtained using LiDAR odometry tightly coupled with simultaneous localization and mapping(SLAM) algorithm

and then noise reduction and dimensionality reduction are carried out to obtain a 2D grid navigation map.Secondly

the re-localization method of adaptive Monte Carlo localization (AMCL) is improved to enhance navigat

ion and localization accuracy.Then

the local path planning algorithms of Safe A

*

and timed elastic band (TEB) are fused to enhance the robot’s obstacle avoidance ability during the inspection process in pipe gallery.Finally

functional verification is carried out in a laboratory simulation scenario.Compared with the original point cloud map

the amount of data is reduced by 88.87%

and the optimized AMCL re-localization position matching error is reduced by 86.88%

and the orientation matching error is reduced by 45.26%.The results show that the navigation system can still realize navigation and localization functions under feature-degraded scenarios

which proves its superior performance in practical applications.