Research On Visual Servo Control Method Of Automatic Drilling Robot

In demolition blasting scenarios,it is necessary to drill blast holes manually.However,manual drilling has problems such as high labor intensity,poor working environment,high dust and noise during construction,and personal safety of construction personnel.As unmanned intelligent robots replace manual operations,they have also become an important goal in the development of demolition blasting.Based on the above background,it is proposed to use an automatic drilling robot to replace manual drilling in demolition blasting.The visual servo control method for drilling robots proposed in this paper is based on the application background of accurate positioning and automatic drilling tasks in demolition blasting,and studies the visual servo control method for automatic drilling robots.The purpose of this paper is to provide a visual servo control method for robotic drilling to achieve automatic drilling of building objects to be blasted.The main research content of this article is as follows:Firstly,On the verification platform of automatic drilling robot technology,the visual servo experimental platform of drilling robot based on UR5 cooperative manipulator is built.The kinematics of the manipulator is modeled,and its forward and inverse kinematics are solved according to the joint coordinate system of the automatic drilling robot.The correctness of the kinematics of the automatic drilling robot is verified by using the robot-toolbox toolbox in MATLAB.The camera’s internal parameters are calibrated for the eye in hand model of an automatic drilling robot using eye in hand.Secondly,For the drilling of column wall in the demolition blasting scene,the normal vector estimation of column wall based on point cloud is proposed.After the point cloud camera collects the wall point cloud data,it carries out the point cloud data preprocessing,cuts and filters out the point cloud noise other than the wall point cloud data,and then carries out the point cloud plane segmentation and plane normal vector estimation.Then select the center point of one of the planes,and combine the plane normal vector to get the axis angle data of the center point.After combination,get the coordinates of the point in Cartesian space,and verify the data accuracy of the plane normal vector estimated by the point cloud through the mechanical arm.Aiming at the drilling task of an entire wall,a visual servo control method for an automatic drilling robot is proposed.Firstly,the robot arm is calibrated with the hand and eye,and then an April tag is used as the target object.Finally,based on the tag as the target object,image based visual servo(IBVS)and position based visual servo(PBVS)are performed for simulation and real robot arm visual servo experiments.The experimental environment considers simulating demolition blasting scenarios.Firstly,error convergence experiments are conducted on IBVS and PBVS for vertical wall drilling operations under good lighting conditions(referred to as the standard operating environment),and the experimental data under this operating environment is used as reference data to compare the results of visual servo task execution after changing the operating environment.Subsequently,orthogonal experiments were conducted on three factors that may affect the robustness of visual servo task completion,namely different lighting conditions,different target object postures,and hand eye calibration accuracy.By analyzing the result data of two kinds of visual servo tasks,it is concluded that the robustness of the control method of automatic drilling robot based on image-based visual servo using point feature is better than that of position-based visual servo control method,the visual error convergence curve of IBVS and the camera motion trajectory are better than those of PBVS,and the accuracy of visual error convergence can meet the drilling requirements of demolition blasting.