Research On Indoor Construction Robot Fusion Localization Based On BIM Information

In recent years,with the advancement of modernization and intelligence in the construction industry,indoor construction robots have gradually replaced manual construction operations.In a complex construction environment,the basic condition for a construction robot to achieve precise construction is that the robot has high-precision positioning information.Traditional indoor robot positioning mainly relies on its own various sensors such as Ultra Wide Band(UWB),Li DAR sensor,vision sensor and inertial sensor(IMU)to achieve robot positioning.However,the UWB system requires external When installing equipment,a single sensor is unstable due to its own accuracy and the influence of the external environment.The traditional indoor positioning method cannot meet the positioning requirements in construction scenarios.Based on the above analysis,in view of the problem that the existing positioning schemes cannot satisfy the precise positioning of robots in the construction scene,this thesis mainly studies the indoor global positioning algorithm based on Building Information Modeling(BIM)and Li DAR,and based on global positioning data and inertial measurement.Inertial Measurement Unit(IMU)robot fusion positioning algorithm.The specific research content of this thesis is as follows:(1)Introduce the information structure of the building information model and its standard IFC data structure,construct the BIM scene according to the test scene,and extract the global point cloud map for the follow-up test of the construction robot;analyze the principle of the lidar point cloud down-sampling,And according to the real-time frame data of lidar,the down-sampling effect is verified by experiments;the principle explanation and effect verification of the point cloud ground segmentation algorithm are carried out.(2)Determine the overall design scheme of the robot,design the software and hardware system of the robot,compare various registration algorithms and finally determine the NDT registration algorithm as the positioning algorithm in this thesis,and build a global positioning module based on lidar.(3)Analyzed the error source of the inertial measurement sensor,established the error model and motion equation according to the error,and established the positioning method of pure inertial navigation;then according to the error state Kalman filter fusion algorithm,a BIM-based IMU/LIDAR loose fusion was constructed The positioning system integrates the global positioning information with pure inertial navigation to avoid the failure of the indoor positioning of the robot due to the instability of the single-sensor positioning method.(4)Design the test plan to verify the effect of point cloud processing and ground point cloud segmentation.Design experiments to investigate the effect of fusion positioning,and analyze and process the test data.Figure [41] Table [9] Reference [57]...