Research On Optical Autonomous Navigation For Approaching Phase Of Small Body Exploration

As a sign of measuring economic and technological strength,deep space exploration has always been an area where major powers are committed to development.As a popular direction for deep space exploration in the new century,the small body exploration such as asteroids and comets must accelerate the pace of research.With the progress of computer and image processing technology,The optical autonomous navigation of small body exploration has rapidly developed and gradually become the direction of future.How to process navigation images and extract useful navigation information is one of the key technologies of optical autonomous navigation.With the support of sub-projects of the National 863 Program(Dark and Small Objects Geometric Feature Extraction and Navigation Information Calculation),this dissertation deeply study of the optical autonomous navigation algorithm based on navigation images for approaching phase of small body exploration.First,combining the navigation background for approaching phase of small body exploration,the relevant coordinate is established.Then,the orbital dynamic model of detector and the target small body is build by analyzing their orbital motion characteristics.On this basis,this paper simplify the model according to analyze the influence of each perturbation force.At the same time,according to the characteristics of optical autonomous navigation,a camera observation model is established,and then,this paper summarize the extraction method of observation information.Second,multiple-points and single-point images are processed separately according to the difficulties of their own images.In the multiple-points image,mean filtering is used to eliminate the noise,and the polynomial fitting method is used to solve the tailing problem.Afterwards,the point target is identified by the trianglematching algorithm,and the observation information is obtained.When processing single point images,this paper focus on solving the problem of dim small target extraction.The maximum inter-class variance method is used to improve the image signal-to-noise ratio,and then combined with an improved mathematical morphology method,the simulation results show that the algorithm can quickly and accurately extract dark and small targets.Next,for the area target image,this paper deals with the image in sequence through astigmatism repair,edge detection and false edge removal.considering the irregular and incomplete contour of the area target,this paper gives the moment algorithm and the least square method.In order to solve the problem that the least squares doesn't consider the differences between sample points,a new method for screening and grouping sample points is designed.The simulation results show that the improved algorithm can accurately and effectively obtain the target celestial contour and centroid information.Finally,this paper presents the optical navigation algorithm for the small body exploration.Point target navigation uses Kalman filtering to predict the movement position of the target small body and reduce the image search range.The simulation results show that the effect of random errors can be suppressed to achieve the tracking of the target point.In terms of area target navigation,an optical navigation algorithm based on centroid and apparent diameter is proposed,which combines the information of apparent diameter and centroid coordinates to derive the relative position of the detector.The feasibility of the algorithm was verified by using the desktop semi-physical simulation system.