Font Size: a A A

Research On AGV Localization For Industrial Applications

Posted on:2022-01-30Degree:MasterType:Thesis
Country:ChinaCandidate:S T YaoFull Text:PDF
GTID:2518306335966909Subject:Control Science and Engineering
Abstract/Summary:PDF Full Text Request
Automated guided vehicles,namely AGVs,have brought great convenience to transporta-tion in many industry fields,such as manufacture,catering service,pharmaceutical industry etc.Many AG V applications in these industry fields have proved that the stability and accuracy of the localization are crucial to the usefulness of the AGVs.To enhance the robustness and impro ve the accuracy of the current localization methods adopted for AGV industry applications,three layers studies have been conducted:1)Considering that most of the AGVs employ the reflector-based localization method in industry applications,a set of optimization strategies have been proposed.These strategies include a corresponding optimization approach,an outlier elimination approach,a laser motion compen-sation approach and a supplementary laser odometry approach.Several experiments carried out in an industry field prove the effectiveness of these proposed approaches.2)Predictably,laser-based SLAM will be widely adopted for AGV localization and navigation in industry applications.Hence,a novel method is proposed for optimization Cartographer algorithm,which is one of the most popular laser-based SLAM algorithms.The proposed method utilizes the intensity information of the laser to improve the accuracy of Loop Closure and further help original Cartographer accelerate the localization processing.Both simulation and experiments prove the valid and effectiveness of the proposed optimization method.3)To reduce the cost of AGVs,this part explores fusing the information from low-cost laser sen-sors with visual information for AGV localization in industry applications.The visual point cloud obtained from RGB-D cameras is transferred to a lidar coordinate system and then the vi-sual point cloud is fused with the laser point cloud generated by low-cost laser sensors.Through this fusing mechanism,the possible missing environment information described by laser point cloud can be compensated by the visual information.Finally,comparison experimental studies are conducted in industrial environment.
Keywords/Search Tags:AGV localization, reflector localization, laser SLAM, camera lidar fusion
PDF Full Text Request
Related items