An Accurate Calibration Method of Large-scale Reference System

The large-scale reference system (LRS) based on multiple measuring stations is widely used to estimate the position and orientation of the aligned components of aircraft. The reference system consists of some enhanced reference system (ERS) points. However, the calibration accuracy of the reference system is prone to be degraded due to measurement systematic errors and environmental disturbance in the workshop. A novel nonlinear optimization method has been developed to reduce the overall calibration error of the LRS. First, the method focuses on analyzing the inhomogeneity and anisotropy of measurement errors of ERS points and establishing an improved weighting model for optimizing the coordinates of the points. Then, the measurement errors of each measuring station can be greatly reduced through a nonlinear unconstraint optimization method with redundant measurements. Thus, a more accurate LRS can be established. Finally, a field measurement experiment of ERS points in the reference system was conducted, and the results showed that the mean registration error of all ERS points was reduced from 0.058 to 0.014 mm and the standard deviation of registration error was reduced from 0.021 to 0.012 mm. Meanwhile, the measurement errors of each measuring station were also greatly reduced.