Error Analysis And Compensation Of Multi Joint Robot

Robots are no longer confined to industrial production,but appear in more application fields.Robotics technology has become an important symbol to measure a country's manufacturing level and technological level.Precision has a great influence on the stability,accuracy and reliability of robot motion,and has become a key index to compare robot performance and measure robot technology.Therefore,it is necessary to study and improve the accuracy of the terminal pose of the robot.In order to improve the motion accuracy of the end-effector of the multi-degree-of-freedom robot,this paper takes the REBot-V-6R robot as an example to carry out the related research.Based on the establishment of the system error model,a new method of position and pose error compensation based on intelligent algorithm is proposed.And the Matlab software and Adams virtual platform co-simulation.The main research contents and conclusions are summarized as follows:1)According to the kinematics principle of the robot,the position and pose of the multi-degree-of-freedom robot are analyzed.The forward kinematics modeling and inverse kinematics solution of the robot were obtained by using the screw theory,and the corresponding nominal kinematics geometric parameters of the robot were deduced.Then the conversion formula of the terminal pose of the robot is obtained.2)On the basis of the kinematics model,three error factors,namely joint axis position error,joint axis direction error and joint Angle error,which have the greatest influence on the pose accuracy of the robot end are added.The system error model is established by the error screw,and the simulation is verified by ADAMS virtual platform,and the validity and accuracy of the error model are confirmed.3)The sensitivity analysis of the errors is carried out,and the errors and joints that have great influence on the accuracy of the end are pointed out,which provide a theoretical basis for the design and assembly of the robot.4)In this paper,the neural network algorithm is used to compensate the error of the robot terminal pose.Firstly,a suitable BP neural network is designed,and then it is improved by using particle swarm optimization(PSO-BP)algorithm.The two algorithms were used for error compensation simulation experiments.After simulation,it can be seen that the predicted error value after BP method compensation is between0.09mm and 0.2mm,and that after PSO-BP method compensation,the predicted error value is between 0.01mm and 0.024mm,which shows that the improved BP neural network has better error compensation effect.Considering the shortcomings of PSO-BP algorithm,such as slow convergence speed,a new swarm intelligent search algorithm(Sparrow algorithm)is proposed to improve BP neural network,namely SSA-BP.After simulation,the compensated error value is between-4.8×10^-3?-3.4×10^-3mm.It can be seen that SSA-BP algorithm has better error compensation effect.