Space Rendezvous And Docking, Optical Measurement Techniques

This paper is arranged under the development of rendezvous and docking (RVD) technologies. The Real-time and accurate measurement of the relative position and the relative attitude between two spacecraft is completed using optical-imaging processor in this paper. On the base of optical measurement theory, a binocular vision model is put forward. The symmetric feature points are configured in the target spacecraft. Under this, the measurement model of motion parameters is erected and discussed. From 100 meters to 2 meters, the restriction of angle of view is considered. During different phase, using different focus of image sensors to keep the image of feature points in the field of view (FOV). And then analyze the three kinds of random noise affecting CCD imaging quality. A straightforward noise statistical model is proposed. A LLMMSE filter is designed to reduce noise based on the noise model to improve SNR of images. Making use of the shape and lightness characters of feature points to detect the feature points from images. Because the accuracy of feature imaging point's centroid has close relationship with measuring accuracy, an algorithm based on iteration of the center of the track window is used to improve the measurement accuracy. The closing velocities between two spacecrafts are very slow during docking. Such a motion is well defined using an appropriate acceleration model. Kalman filtering techniques can thus be utilized to improve tracking accuracy. In the simulating experiment of RVD, a optical measurement system based on DSP is designed. The main modules and the main features of TMS320C6000 are introduced in this paper. At last the experiment at calibrating room verify this system.