Six-degree-of-freedom Industrial Robot End Error Compensation Method

As the representative of advanced electromechanical integration,6-DOF serial robots are widely used in industrial manufacturing.The robot accuracy,including absolute positioning accuracy and repetitive positioning accuracy,is one of the most important indexes to measure the performance of industrial robots.The repetitive positioning accuracy has already reached a higher level,generally in the order of 0.1mm,but the absolute positioning accuracy is difficult to meet the practical needs of industry.Therefore,how to improve the absolute positioning accuracy of robot plays an important role in robot calibration technology.In this paper,the compensation method of robot accuracy is studied in detail by using the KUKA robot in kinematics modeling,error analysis,parameter identification and error compensation,and the experimental verification is carried out.The main research contents and conclusions are summarized as follows:1)According to the principle of industrial robot kinematics,according to the kinematics principle of robot,the pose and precision of six degree of freedom articulated robot are analyzed.KUKA robot kinematics model is established by DH modeling methods,and the robot is deduced in the name of the corresponding kinematic geometric parameters,and get the end-effector pose transformation formula.2)On the basis of kinematic model,the geometric error factors which affect the positioning accuracy of industrial robot are considered.The geometric deviation model is established by differential method,and the model is verified by simulation,and the validity and accuracy of the model are confirmed.3)In the research of industrial robot positioning error compensation method,firstly the geometric error model of industrial robot was used,and the N-R iterative algorithm was used to verify the simulation.The limitation of positioning error compensation is considered by using error model,firstly a suitable BP neural network is designed and then it is improved,namely PSO-BP algorithm,compared with the error model of KUKA robot,the simulation is carried out by network approximation and error compensation is carried out,which proves the accuracy of the improved method.Comparing the two simulation results show that the compensation accuracy of positioning error compensation based on the error of positioning error compensation model based on PSO-BP neural network is higher.4)Finally,taking into account the slow convergence rate of PSO in traditional PSO-BP neural networks,etc.An improved PSO algorithm for PSO algorithm is proposed,and the superiority of that is proved by four classical functions,and the advantages of improved PSO are proved by using four classical functions.Use improved PSO and the BP neural network to make the accuracy of KUKA robot positioning better.