| With the rapid development of modern industry,the application area/field are becoming wider and wider,which results in the higher requirement of the positional accuracy for the end of industrial robots.There are always errors between the actual and theoretical positions for the influence of various error factors,which limits the applications of robots in high precision occasions.Calibration is an important way to improve the positional accuracy of industrial robots.The calibration and error compensation methods for 6-DOF industrial robots are studied in this paper,in which one industrial robot is used as the experimental object.The main works are as follows.According to the mechanical structural characteristics of the industrial robot,the kinematic model is established with DH method.Then,the relationship between the position of the end and the DH parameters is deduced.And the kinematic model is verified by simulation and experiments.The error model of the end position of therobot is obtained through deducing the transformation of DH parameters' errors and positional errors of the end.The kinematics DH parameters of the linear correlation are analyzed,and the impact of the linear correlation on the identification results of the kinematics parameters is presented.Based on robot kinematics model and error model,the least-squares scheme is proposed to analyze the kinematics model of the 6-DOF industrial robot.In order to simplify the step about kinematics parameter identification,a genetic-tuba search algorithm is proposed,and the flow chart of the recognition algorithm is given.The algorithm does not need to analyze and transform the error model,but regards the parameter error as an optimization problem.By using the genetic algorithm to execute global search and the tabu search algorithm to local search,the fitness function value is gained in the end.Based on Robotics toolbox and MATLAB GUI interface,the software about parameter identification is compiled,and simulation identification is executed for industrial robot simulation by using the least-square scheme and genetic-tuba search algorithm respectively,and the minimum amount of data what two kinematic parameter identification algorithm need is identified.The simulation result indicates that the position precision of the end-effector is improved after doing parameters identification and error compensation.The maximum direction error after parameter identification based on the least squares method is from 3mm to 0.005mm.Based on the genetic tabu search algorithm parameter identification,the maximum direction error drops to 0.008mm.Taking coordinate measuring machine as a measuring tool,the relate experiment research is executed.Firstly,the end position of the robot is measured in measuring machine coordinate space.Secondly,the end position of the robot is transformed into the robot coordinate space,and two identification algorithms are used to do kinematic parameter identification separately.Finally,two experimental results are compared and analyzed.In general,the absolute position accuracy of the robot end is usually within 5mm.After identification,the maximum direction error is reduced from 15mm to 0.7mm and 1.4mm,which can meet the requirements of the robot in general.Finally,the experimental results are carried out.Compared with the least squares method,the parameter identification algorithm based on the genetic tabu search algorithm is simple and more efficient,and can focus on the optimization of the algorithm. |
PDF Full Text Request