| With the development of Advanced Manufacturing Technology,the Online Measurement Technology oriented to the manufacturing site has been widely applied.Among them,the robot flexible online measurement technology based on linear structured light has the advantages of non-contact,high precision and high flexibility,which is well suited to the needs of industrial measurement and has become an important research field of the Online Measurement Technology.Robot flexible online measuring system based on linear structured light is composed of robot and linear structured light sensor.The working principle is as follows: the robot drives the sensor to collect the shape information of the object from different perspectives,and then converts the shape information to the robot base coordinate system for splicing based on the Hand-Eye relationship and the robot end pose,so as to realize measurement of the topography of the object.In order to realize the measurement function,it is necessary to calibrate the parameters of the linear structured light sensor and the Hand-Eye relationship.In addition,it is also necessary to calibrate the kinematic parameters of the robot to improve the absolute positioning accuracy of the robot(the deviation between the actual pose of the robot and the expected pose of the robot control instruction).In all calibration tasks,the calibration of robot kinematics and linear structured light sensor are independent,while Hand-Eye calibration depends on the above calibration results.Therefore,this paper adopts the following ideas to carry out the calibration work.(1)Firstly,kinematic calibration of the robot was carried out.The D-H model and position error model of the robot were firstly established(the error of base coordinate system was taken into account),then the Levenberg-Marquardt algorithm was used for parameter identification,finally the cartesian space compensation method was used for error compensation,and the absolute pose accuracy of the robot was improved by 74.3%.(2)Then the calibration of the linear structured light sensor was carried out.Firstly,the camera were calibrated based on Zhang's method,and then the optical plane was fitted based on the principle of ray intersection.Meanwhile,the image information used for camera calibration and optical plane calibration was both collected by changing the sensor pose while keeping the target fixed,which effectively improved the calibration efficiency.(3)Base on this,the Hand-Eye calibration is carried out.Firstly,a nonlinear optimization model of Hand-Eye relationship is established based on spatial plane constraints.Then,the Augmented Lagrange Multiplier method and Levenberg-Marquardt algorithm are used to solve the model to realize synchronous calibration of Hand-Eye pose relationship.(4)Finally,the measurement experiment of standard parts is carried out.By comparing the measurement results of the system with those of the laser tracker,it can be concluded that the distance measurement accuracy of the system is 0.076 mm,the plane measurement accuracy is 0.1075 mm,and the three-dimensional measurement accuracy(the standard deviation of point cloud deviation from the true value)is 0.076 mm. |
PDF Full Text Request