Observability index optimization of robot calibration based on distance error model

Configuration optimization based on the observability index is an important method to improve the accuracy of robot calibration. However, due to the lack of position information, the traditional observability index is unsuitable for robot configuration optimization based on the distance error model. Therefore, this paper proposes a distance observability index based on a distance error model for robot calibration. Firstly, the MD-H method is employed to establish the kinematic model of the general six-degree-of-freedom (six-DOF) industrial robot and derive an equation combining distance error and joint error. Then, the distance observability index O d1 is constructed based on the distance error model and Fisher Information. Finally, the data set is constructed in the robot workspace, and the DETMAX algorithm is used to select 50 configurations with a high observability index as the optimal configuration set. The kinematic parameter error of industrial robots is identified by the optimal configuration set. The experimental results show that the proposed observability index optimization improves the data set’s observability from 0.234 to 0.331, reduces the industrial robot’s max distance error from 0.464 mm to 0.340 mm, mean error from 0.127 mm to 0.095 mm, and RMSE from 0.296 mm to 0.218 mm, which verifies the efficiency of the proposed method.