Positioning Accuracy Reliability Analysis And Error Compensation Of Industrial Robot

To improve the productivity and accuracy of repeated work,industrial robots have been widely used in various fields of industrial production including installation,handing,production lines and fashioning at present.However,due to the existence of the uncertain variables such as joint clearance,axis length machining error,and installation error of the industrial robot,the actual movement position of the end-effector of industrial robot deviates from the target position which is called positional error.Positioning accuracy is an important performance index of industrial robot,which directly affects the quality of processing parts and assembly parts of industrial robot.Positioning accuracy mainly includes two aspects:repeated positioning accuracy and absolute positioning accuracy.At present,the absolute positioning accuracy of industrial robots has not met some finishing requirements.Therefore,the key to improve the performance of industrial robots is to accurately analyze the positioning accuracy reliability of industrial robots and improve the absolute positioning accuracy.This paper analyzes the positioning accuracy reliability and compensate the positional error using radial basis function network.The main research contents are as follows:(1)The classical D-H modeling method is used to establish the kinematics equations of industrial robots and the forward and inverse kinematics are solved which is verified by robot toolbox in MATLAB software.(2)The uncertain variables that affect the positioning accuracy of industrial robots are substituted into the kinematic model and the radial basis neural network algorithm is used to analyze the kinematic reliability.In view of the shortcomings of general neural network learning algorithm,a hybrid learning algorithm is proposed to train radial basis neural network,and the advantages of self-organizing center selection algorithm and orthogonal least square algorithm are combined to improve the computational efficiency on the basis of ensuring the computational accuracy.(3)The radial basis function neural network optimized by particle swarm optimization algorithm is used to construct the mapping relation between the nominal position and the positioning error of industrial robots.A positioning error field concept is proposed to describe the positioning error situation in a specific workspace of an industrial robot.The nominal position is pre-biased to compensate for the error according to information of error field and the validity of the method is verified by a simulation model.(4)According to the proposed radial basis neural network error compensation method,the error compensation experiment is carried out on the actual industrial robot.Three kinds of industrial robots with different rated loads are selected for verification in the experiment.After compensation by the proposed method,the absolute positioning accuracy of the three kinds of industrial robots is significantly improved and the range of positioning error is significantly reduced.Through theory,simulation and experiment,the positioning accuracy reliability analysis and positioning error compensation of industrial robots are carried out.The results show that the proposed method can accurately analyze the positioning accuracy reliability of industrial robot and greatly improve their absolute positioning accuracy,which has important engineering significance.