Rapid, high-robustness and autonomous fixed star identification method

Abstract

The invention provides a rapid, high-robustness and autonomous fixed star identification method, which comprises the following steps: firstly, finding a navigation star pair with a highest matching frequency from all observed star pairs to generate a matching matrix; secondly, confirming the matching relation between every two matching pairs in the matching matrix to generate a matching confirmation matrix; finally, calculating a maximum matching group according to the matching confirmation matrix to obtain an identification result. The method of the invention has the advantages of high identification speed, good robustness, high identification probability, low mis-identification rate, low algorithm realization cost, and high identification error tolerance rate.

Description

technical field [0001] The invention relates to a fast and highly robust autonomous star recognition method, which belongs to the technical field of star map recognition of star sensors. Background technique [0002] The star map recognition algorithm is one of the core technologies of the star sensor. It refers to extracting the observed star and its characteristics to be recognized from the star map captured by the star sensor without knowing the direction of the optical axis of the star sensor ( The most commonly used features are the magnitude of the stars, the angular distance between the stars, the geometric shape composed of multiple observation stars, etc.), and compared with the navigation star library of the star sensor, the corresponding relationship between the observation star and the navigation star is established , to prepare for the attitude calculation of the star sensor. Star map recognition can be divided into all-sky star map recognition and local star m...

AI Technical Summary

Problems solved by technology

misidentification

[0006] The raster algorithm is a typical star map recognition algorithm using a pattern matching strategy. It has the characteristics of small storage capacity, high success rate, good real-time performance and robustness. There is no need to use the advantages of star brightness characteristics, etc., but the main disadvantage is: after the main star is selected, the establishment of the observation mode depends entirely on the reference star. If different reference stars are selected, the established observation modes are completely different, and there is no relationship between them. Contact, if the correct reference star cannot be selected, the corresponding observation mode and navigation mode will not match

[0009] The advantages of this method are: high recognition probability, low cost of algorithm implementation, and high recognition error tolerance rate. reference star, the observation template and the navigation template will not match, and this identification method needs to establish a polar coordinate template for each navigation star in the navigation star catalog. For small star sensors with small field of view and high sensitivity, the navigation star catalog The number is large, and the polar coordinate template of the navigation star requires a large storage space. At the same time, the observation template needs to be matched with each navigation template once, which requires a long recognition time and poor real-time performance

[0010] Among the above recognition methods, the triangle method and the matching group method require a lot of preparatory work to ensure the success rate of star recognition, and the cost of algorithm implementation is relatively high

On the other hand, because the matching group method and the triangle method only use the star-to-star angular distance (side length) to describe the geometric characteristics of the navigation constellation and the observation constellation, it is impossible to prevent the occurrence of "misidentification of the mirror image" in principle; at the same time, the grid The grid algorithm, matching group method and polar coordinate template matching method need to establish a matching template in advance, which requires a large storage space, and after the main star is selected, the establishment of the observation mode depends entirely on the reference star or reference star. If the correct reference star cannot be selected star or reference star, the corresponding observation mode and navigation mode will not match; the neural network recognition algorithm requires a large training set, the real-time performance is poor, and the algorithm has relatively high requirements for hardware

Especially for small star sensors with small field of view and high sensitivity, the number of navigation star catalogs is large, and the grid algorithm, matching group method and polar coordinate template matching method all need to establish a template in advance, which requires a large storage space and takes a long time for template matching. Poor real-time performance

Images