Research On Orbital Calibration Technology Of Elements Of Interior Orientation In Space-based Large Field Of View Telescope

Space has become a fierce competition in various countries.In order to cope with the threat from space,we must comprehensively monitor all on-orbit targets.The space-based large-field telescope plays an indispensable role as an important load for space surveillance.Due to the impact and vibration during launch,and the influence of the harsh space environment during the orbit,the internal orientation elements of the space-based large-field telescope will change after the orbit with respect to the ground calibration.It may affect the accuracy of the on-orbit measurement.Therefore,it is of great significance for the on-orbit calibration of the orientation elements in the space-based large field of view telescope.This paper will explore the on-orbit calibration of the orientation elements in the space-based large field of view telescope.The main research is to analyze the sub-pixel subdivision and star map simulation technology of star point,the star map preprocessing and the calibration process of the calibration star point.And an on-orbit element on-orbit calibration algorithm.In this paper,the working principle of space-based large field of view telescope is analyzed by investigating domestic and foreign literatures.On this basis,the on-orbit measurement model and on-orbit calibration model of space-based large field of view telescope are established.A high-precision simulation of the star-shaped image of the space-based large-field telescope was carried out.A simulated star-point subdivision method based on Gaussian distribution was used,and the sub-pixel gray-scale correction was performed on the energy center of the star point.The simulation of the star map was completed in conjunction with the established star library.In the aspect of centroid point centroid extraction,this paper firstly preprocesses the noise star map,including filtering denoising,threshold segmentation and connected domain partitioning.After analyzing three centroid extraction algorithms and simulating the deviation of the extraction results,the method of extracting the smallest error is selected.Aiming at the large number of star points in the field of view of the space-based large field of view telescope,the appropriate calibration star points are screened by comparing the average deviation of the star point centroid extraction under different conditions,so that the calibration result can be more accurate.Finally,the on-orbit element calibration algorithm is studied and analyzed,and the simulation experiment is carried out.The deviation between the calibration result and the simulated true value is calculated to measure the accuracy of the calibration result,and the average inter-satellite angular distance is used.Statistical bias is used to evaluate the calibration results.Aiming at the uneven distribution of randomly measured noise,the erroneous selection of the star point is proposed.A method of on-orbit calibration of the inner azimuth element with weighted culling is proposed.This method selects the reasonable weight to carry out the bad points with large random noise.Weighting processing reduces or even eliminates the weight of dead pixels.The simulation results of the algorithm show that after using this method to eliminate the bad points,the calibration results are significantly improved compared with the extended Kalman filter algorithm.In the case of focal plane distortion in the image,the distortion model is established and the influence of the focal length under the first-order radial distortion is analyzed.A distortion compensation model is established to compensate for the focal plane distortion existing in the star point.Through simulation analysis,the focal length is accurately calibrated under this model.