| Industrial 6-axis robots belong to a multi-joint series mechanism.Due to errors in the production,manufacturing,and assembly processes,their absolute positioning accuracy is poor,and it is difficult to meet increasingly more demanding work scenarios.Therefore,errors of industrial robots need to be compensated.Make the robot's actual arrival position consistent with the ideal arrival position.This paper proposes a two-stage error compensation scheme for kinematic geometric and non-geometric parameter errors that affect the positioning accuracy of robots.First,it compensates for geometrical parameter errors in kinematics.Then a mesh is created in the joint space to compensate for nongeometric parameter errors.Finally,experiments prove that the above scheme can effectively improve the absolute positioning accuracy of the robot.The main tasks accomplished in this paper are:The kinematics basis and error analysis of the robot: Using the HSR-JR605 robot as an example,the kinematics DH model was established and the kinematics forward solution was solved.The errors affecting the robot positioning accuracy are divided into kinematic geometric parameters and non-geometric parameters.For the two types of error factors,an experimental scheme for the two-level layer error compensation is proposed.(2)Kinematics geometric parameter error calibration: First,the source of the geometric parameter error was analyzed.Then the main flow of error calibration of the kinematics geometric parameters was introduced,including the establishment of position error model,measurement,parameter identification and error compensation.A pose-optimization strategy is proposed for the Jacobian matrix condition number that influences the stability of the robot model parameter identification,and a pose-optimized MATLAB-GUI program suitable for HUASHU robots is written.In order to verify the influence of Jacobi matrix condition number on the parameter identification result,the kinematics parameter error calibration simulation program was programmed on MATLAB,and two sets of different poses(a set of optimized attitude and a set of unoptimized attitude)were simulated and compared.The results show that the calibration compensation results in the optimized attitude are significantly better than the unoptimized attitude compensation results.(3)Joint space gridding error calibration of non-geometric parameters error: The study found that it is not sufficient to compensate for the robot's kinematic geometry parameter error,and the influence of robot's non-geometric parameters on the positioning accuracy of the robot can not be ignored.Because the sources of non-geometric parameters are numerous and they are coupled with each other seriously,it is difficult to perform independent modeling analysis.Therefore,compensation measures for errors in nongeometric parameters used to create meshes in the joint space are used.The principle of grid space error compensation in joint space,the grid division strategy and the compensation method are introduced.(4)Experimental verification: The laser tracer was selected as the measuring instrument,and the HSR-JR605 robot was used as the experimental object.The experimental scheme was designed and the two-level calibration scheme of the robot calibration was experimentally verified.That is,the kinematics geometric parameter calibration experiment was performed first.Based on this,the non-geometric parameter error calibration experiment of joint space gridding was performed.After two-step compensation experiment,the average positioning error of the robot was increased from 4.1113 mm to 1.0608 mm,and the error was improved by 74.2%. |
PDF Full Text Request