Stiffness Parameter Identification And Error Compensation Of 6R Industrial Robot

In aerospace and other fields,the demand for robot automatic hole making of curved parts and medium and large parts is becoming increasingly strong.However,the robot cannot meet the requirements of positioning accuracy of automatic hole making,it becomes the bottleneck of hole making technology.How to improve the absolute positioning accuracy of industrial robots has become a key core technology.Therefore,based on the geometrical parameter identification,the robot joint stiffness identification and joint stiffness error compensation are investigated.The excellent results are verified by selecting sampling points in the robot’s working space and by positioning the aircraft cockpit cover for linear and curved hole making.The experiment shows that the absolute positioning accuracy of the machine can be improved by the joint stiffness error compensation on the basis of geometric error compensation,which can meet the engineering requirements.Specific research contents are as follows:(1)The kinematics of the robot is analyzed,and the D-H motion model of KUKA KR16-2 robot is established through homogeneous transformation matrix,and its correctness is verified.According to the vector product method and virtual work principle method,the robot velocity and force Jacobian matrix is constructed,which is the theoretical basis for the following chapters.(2)The robot joint stiffness model is established to identify robot joint stiffness parameters.The mapping relationship between joint stiffness and Cartesian stiffness of robot is established by using mechanical knowledge and Jacobi theory to provide theoretical basis for the identification of joint stiffness of robot.The influence of complementary stiffness on the stiffness mapping relationship is considered,and both are investigated by means of robot dexterity metrics and quantitative analysis.The optimal machining performance region of the robot is also obtained as well as improving the accuracy of the joint stiffness parameter identification.(3)An experimental platform is built to identify robot joint stiffness parameters and conduct joint stiffness error compensation experiments.The experimental platform is built with KR16-2 robot,laser tracker and 16 kg disk-shaped weight.On the basis of geometric parameter error compensation,sampling points with good dexterity are selected and the changes in the position of the end before and after the end load were measured by the laser tracker.The stiffness parameters are identified and verified by the stiffness mapping relationship.(4)The improvement of robot positioning accuracy is achieved by means of a joint stiffness error compensation strategy and error prediction compensation using an optimised neural network.The stiffness error of the robot joint is a non-geometric error and cannot be described by a fixed mathematical formula.According to the motion characteristics of the robot,based on the error identification of geometric parameters,the joint space grid division compensation strategy is used to compensate the positioning error of the robot end.In order to verify the effectiveness of the compensation strategy,the plane seat cover is positioned by linear hole making and curved hole making.The experimental results show that the joint stiffness error compensation based on geometric parameter error compensation can improve the absolute positioning accuracy of the robot and meet the engineering requirements.