Analysis And Compensation Research On Elastic Error Of Series Six Degrees Of Freedom Robot

Accuracy is one of the important indicators for evaluating the performance of robots.When the robot is subjected to external loads,inertial forces,etc.during its work,the member deforms elastically,causing the actual posture to deviate from the theoretical posture,affecting its end accuracy.Series robots are widely used because of their large working space and flexible joints,but compared with parallel robots,they have weaker stiffness and are prone to elastic deformation,which seriously affects the accuracy of the end actuator.Therefore,it is necessary to study the elastic error of the serial robot in depth.Aiming at the problem that the elastic deformation of the tandem robot subjected to external load under the quasi-static state affects the accuracy of the end,in this paper,a special experimental end effector six degree-of-freedom robot in series is taken as the object.The six-degree-of-freedom robot has carried out in-depth research on elastic error analysis and compensation.Because the robot generally works at a low speed,it ignores the inertial force it receives and only considers its own gravity and external load.An elastic error analysis and compensation method for tandem robots under quasi-static is proposed to improve the accuracy of the end pose of the tandem robots.The work carried out in this article is as follows:(1)Study the kinematics of the six-degree of freedom robot,Analyze the kinematics forward and inverse solutions of the linear variable arc mechanism in both yaw and pitch directions.Establish the direction cosine of each sub-robot,derive the forward and backward solution formula of the end position of the robot,and draw the conclusion that six degrees of freedom are completely decoupled at the end of the robot.(2)Analyze and compare the existing static stiffness solving methods of the robot,combine the characteristics of the six-degree-of-freedom robot,and use the finite element method to analyze the static stiffness of the robot.The overall static analysis and weak parts of the robot are analyzed,and a series of coordinate transformations are used to solve the six degrees of freedom direction deformation at the end of the six degrees of freedom robot.The components with large deformation are considered as flexible bodies,and the remaining components are considered as rigid bodies.With the help of SOLIDWORKS and ANSYS joint batch simulation,the problem of re-meshing when changing the pose in the finite element method simulation is solved,and the stiffness coefficient matrix under each pose is obtained.(3)The stiffness coefficient matrix fitting prediction method is studied.Analyze the characteristics of the samples,respectively use BP neural network and deep confidence network to fit the stiffness coefficient matrix,establish two models of fitting prediction method,and compare the prediction results of the two methods.(4)The inverse kinematics compensation method is used to compensate the elasticity error of the six-degree-of-freedom robot.In order to verify the correctness of the elastic error and posture compensation,randomly select multiple sets of postures to analyze the elastic error under a certain load,and perform posture compensation through the positive kinematics compensation method.The actual posture of the compensated robot is closer to the theoretical posture,and the posture error caused by elastic deformation is greatly reduced,which proves that the method is effective.