Robot Calibration Based on Improved Exponential Optimization with Iterative Weighted LM Method
Aiming at the cumbersome and inefficient factory calibration process of collaborative robots,this paper proposes a kinematic calibration algorithm based on a two-step strategy,which combines the im-proved exponential optimization algorithm with the iterative weighted LM algorithm to simplify the calibra-tion process and improve the positioning accuracy. First,a robot parameter identification model is estab-lished based on the improved DH method and position differential error transformation,and a calibration system error model is built by combining the coordinate system of the measurement device. The improved exponential optimization algorithm is used to quickly obtain the initial parameters of the measurement coor-dinate system. Secondly,in order to improve the robustness of the identification results,the distance residual is used as a weighting factor,and the geometric parameter error of the robot model and the matrix parame-ter error of the measurement coordinate system are compensated by the iterative weighted LM algorithm. Fi-nally,through experimental validation,the results show that the mean squared error and maximum error of the robot position error are reduced by 80.28％,71.61％ and 52.16％,respectively,which verifies the cor-rectness and effectiveness of the method.
万方数据
维普
