Speaker
Description
This study addresses the core problem of how to achieve high-precision measurement consistency in large-scale engineering metrology under high-noise and multi-disturbance environments.
First, for multi-station laser tracking systems, a GUM-based uncertainty propagation model is established, revealing the evolution laws of measurement accuracy and enabling accuracy optimization.
Second, to address the issues of low efficiency and poor adaptability in node layout for reference measurement, a node weight evaluation and optimization method based on network analysis is proposed. This method constructs node geometric relationships using 3D Delaunay tetrahedral networks and computes the total weight of each node based on barycentric coordinate invariance to guide node selection and network simplification. It exhibits strong robustness against variations in survey stations, measurement errors, and scale factors. In the application on the Hefei Advanced Light Facility platform, the number of nodes is reduced by 31%, while the calibration accuracy remains at 9.8 micrometers.
Finally, to address the coupling between registration and deformation in structural deformation measurement, an error propagation model for point cloud registration residuals is established, revealing the mechanisms by which deformation and noise affect rotation, translation, and residual distribution. A linear deformation inversion method is then proposed. Compared with conventional methods, the deformation estimation accuracy is improved by 46% to 97%, and the method remains stable across different deformation fields. Furthermore, the coordinate transformation accuracy is enhanced through corresponding compensation methods.
