The attitude control error of the robot end-effector directly affects the manufacturing accuracy. The study aims to develop a real-time measurement method of the industrial robot end-effector attitude in the field environment for improving the control accuracy of robot attitude.

In this paper, an attitude measurement method of robot end-effector based on the inertial technology was proposed. First, an inertial measurement system was designed, and the measurement parameters and installation errors were calibrated. Then the inertia measurement principle of robot end-effector attitude was explored, and the robot end-effector attitude measurement was realized with the fourth-order Runge−Kutta algorithm. In addition, the influence of the data processing algorithm and sampling frequency on the attitude accuracy was analyzed. Finally, a test platform was built to experimentally explore the proposed inertial measurement method.

The inertial measured data were compared with the data obtained with the laser tracker. The measurement accuracy of the inertial measurement method reached 0.15°, which met the accuracy requirements of real-time measurements of robot end-effector attitude in the manufacturing field.

The method proposed in this paper is convenient and can realize the real-time attitude measurement of industrial robot. The measurement results can compensate the attitude control error of the robot end-effector and improve the attitude control accuracy of the robot.

One critical issue in network real-time kinematic (NRTK) is the interpolation of atmospheric delay for user stations. Some classic interpolation algorithms, such as linear interpolation method (LIM), ignore the strong correlation between tropospheric delay and height factors, and the interpolation accuracy is poor in areas with large height difference. To solve this problem, a troposphere modelling method based on error compensation, namely ECDIM (Error Compensation-Based DIM), is proposed, and this method can be applied to both conventional single Delaunay triangulated network (DTN) and multi-station scenarios. The results of California Real Time Network (CRTN) with large height difference show that compared with LIM, the overall modelling accuracy with ECDIM has been improved by 50.1% to 67.3%, and especially for low elevation satellites (e.g., 10–20 degree), the accuracy is increased from tens of centimetres to a few centimetres. At user end, the positioning error in up direction with LIM has an obvious systematic deviation, and the fix rate of epoch is relatively low. This situation has been improved significantly after using ECDIM. The results of Tianjin Continuously Operating Reference System (TJCORS) show that in areas with small height difference, both methods have achieved high precision interpolation accuracy, and the positioning accuracy with ECDIM in up direction is improved by 21.2% compared with LIM.