Research On Position And Attitude Measurement System Based On Rendezvous And Docking Simulation Platform

Rendezvous and docking (RVD) process is a very perplexed technology, and measurement technology is the key section in RVD. With the development of technologies in RVD, the relative position and attitude measurement of spacecraft is very significant. In this dissertation, the key technologies of camera calibration, image feature extraction and measurement of the relative position and attitude have been investigated. Meanwhile, design and implementation of measurement system for the relative position and attitude of spacecraft have been done. According to experimental data detected from the simulation platform, this system has been proved to be feasible.The models of camera and imaging in machine vision have been studied, in this thesis. Based on the imaging principle of CCD camera, it is described that camera calibration is central in the vision-based measurement. Thus, a stable and flexible Zhang-calibration algorithm has been introduced in detail, and camera has been calibrated using templates. On the results of the calibrating simulation, the selected calibration algorithm is better and has a good performance. Then it provides precision foundation for the subsequent processes of feature extraction and coordinate estimation in the whole measurement system.In consideration of actual application, the theory and algorithms of seed filling method, feature point coordinates extraction, monocular-binocular position and attitude measurement are well systematically studied. Finally, a more rapid improved seed filling algorithm and a POSIT algorithm with less feature points is determined in vision-based position and attitude measurement system, to ensure precision and efficiency of the calculation. On the platforms of laboratory, software and hardware of the entire measurement system is designed and built. Considering the situation that monocular field angle is limited and the target might be obscured, the calculation algorithm blending monocular and binocular is developed including scheme of IMU (Inertial Measurement Unit). In the actual experiment simulation platform,it has achieved many nice results.