Development Of A Wall-climbing Robot For Large Steel Structure Inspection

Large steel structure equipments,such as large oil storage tanks,wind turbine towers,ships,etc.,have been exposed to harsh natural environments for a long time,and the surfaces are prone to rust and cracks.Secondly,mechanical equipments such as cranes have high operating intensity,fatigue failure,structural fracture and other failures often occur.At present,the detection of these defects mainly relies on manual work,and the labor intensity and risk factor are extremely high.Therefore,it is urgent to develop wall-climbing robots to replace manual inspection operations on large steel structures.The height of large steel structures is up to tens or even hundreds of meters,so the wallclimbing robot is required to have a remote control function in terms of control.In addition,the outer surface area of steel structures such as ship outer plates is large,so the manual remote control requires a long time operation,and the work efficiency is low.The autonomous movement of wallclimbing robots on such walls can further liberate the labor force.Therefore,this paper proposes to apply SLAM technology to wall-climbing robots,so that the robots have the ability to work autonomously and improve their intelligence.The main research contents are as follows:(1)Structure design and force analysis of the wall-climbing robot.According to the actual working conditions of large steel structure inspection,the permanent magnet adsorption and wheeled movement schemes are analyzed and adopted,and then the wall-climbing robot body structure is designed.Aiming at the instability forms of the robot: sliding along the wall and overturning longitudinally and laterally,the spatial pose model and static model of the robot are established to solve the adsorption force under reliable adsorption.In order to meet the requirements of the robot to move flexibly,the force analysis of the motion state is carried out to make the robot’s own gravity,motor torque and adsorption force match each other.(2)Magnetic field analysis and optimization design of permanent magnet wheel.Materials of the permanent magnet wheel are selected,and magnetic circuits of the permanent magnet wheel are designed according to the magnetizing direction of permanent magnets and the arrangement of the magnetic materials.Based on Maxwell electromagnetic field theory,the COMSOL software is used to analyze the magnetic field characteristics of different magnetic circuits,and the more advantageous permanent magnet wheel magnetic circuit is selected.In order to improve the utilization of magnetic energy,the influence of structure parameters on the adsorption force and efficiency of the permanent magnet wheel is analyzed,and the optimal structural size is determined.After optimization,the adsorption force of the permanent magnet wheel reaches 241.03 N,the adsorption efficiency is increased by 20.5%,and the robot has the ability to adsorb on the large curved steel structure reliably.The test results show that the robot can adsorb stably on the steel wall,verifying the feasibility and correctness of the optimized design of the permanent magnet wheel.(3)Design of remote control system for the wall-climbing robot.On the basis of the prototype manufacturing,the two-level distributed remote control system of handheld remote control terminal and robot vehicle terminal is designed,and the hardware circuit design and selection of the system are completed.The Open Wrt system in Wi Fi module is configured and developed,and thr Android client is used as a handheld remote control terminal to realize command transmission based on TCP protocol,Socket and video transmission based on HTTP protocol.Finally,the program design of the main control chip of the robot vehicle terminal is carried out.The test results show that the handheld remote control terminal can accurately control the actions of the robot at a distance of about 100 meters,and the robot can move flexibly on the steel wall.The video transmission delay is 0.22s～0.45 s,and the video image is clear and the transmission is stable at a distance of about 100 meters,which basically meets the needs of large steel structure inspection.(4)Research on autonomous movement of the wall-climbing robot.The robot motion model,track estimation model and lidar observation model are established,and two kinds of SLAM algorithms,Gmapping based on RBPF and Cartographer based on graph optimization are studied.Based on the move＿base navigation framework,the AMCL positioning algorithm,D*Lite global path planning algorithm and DWA local path planning algorithm are used to realize the robot’s autonomous navigation function.For the robot autonomous mobile platform,the related hardware circuit design and selection are carried out,and the development of chassis drive and ROS system node program is completed.Set up experimental scenes for SLAM mapping.The results show that the relative error of Cartographer algorithm mapping is between 0.24% and 2.67%,which is more accurate and robust than Gmapping algorithm.The robot navigation test is carried out to verify the robot’s path planning ability,as well as the feasibility and rationality of combining SLAM technology with wall-climbing robots for large steel structure inspection.