Starlight navigation device and positioning navigation method thereof

Abstract

The invention relates to the technical field of aerospace navigation, in particular to a starlight navigation device and a positioning navigation method thereof. The light shield is used for filtering light pollution generated by the aircraft; the lens is used for acquiring starry sky picture information, the light shield is connected with the lens, and the light shield is arranged at the front end of the lens; the calculation circuit is used for receiving the starry sky picture information and calculating to obtain a feature vector of the starry sky picture information; and the star catalogue memory is used for receiving the feature vectors, matching constellation information corresponding to the feature vectors and returning the constellation information to the calculation circuit, and the calculation circuit sends the constellation information to the aircraft and calculates position information. The starlight navigation device comprises a light shield, a lens, a calculation circuit and a star catalogue memory, and the angle of the aircraft relative to an observation star can be calculated according to the starlight vector measured by the starlight navigation device, so that the attitude angle of the aircraft is calculated, and starlight navigation is realized.

Description

Technical Field

[0001] The present invention relates to the field of aerospace navigation technology, and in particular to a starlight navigation device and a positioning and navigation method thereof. Background Art

[0002] When flying outside the atmosphere, the stars around the earth can be clearly identified because there is no atmospheric interference. With the help of starlight sensors, the identified starry sky can be matched one by one with the starry sky pictures in the storage device, and the current coordinates can be calculated based on the relative positions of key stars.

[0003] Traditional GPS methods cannot provide location services for objects moving at ultra-high speeds, and fiber-optic guidance is easily affected by external light environments, such as satellite reflections, cloudy weather, and other external conditions that affect optical imaging results. Summary of the invention

[0004] In view of the above-mentioned shortcomings of the prior art, an object of the present invention is to provide a starlight navigation device and a positioning and navigation method thereof to solve the problem that the prior art cannot provide location services for objects moving at ultra-high speeds and is easily affected by the external light environment.

[0005] To achieve the above-mentioned and other related purposes, the present invention provides a starlight navigation device, comprising:

[0006] Sunshade, used to filter light pollution generated by aircraft;

[0007] A lens, used for acquiring starry sky picture information, wherein the light shield is connected to the lens and is arranged at the front end of the lens;

[0008] A calculation circuit, used for receiving the starry sky picture information and calculating a feature vector of the starry sky picture information;

[0009] The star catalog memory is used to receive the feature vector, match the constellation information corresponding to the feature vector, return the constellation information to the calculation circuit, and the calculation circuit sends the constellation information to the aircraft to calculate the position information.

[0010] In one embodiment of the present invention, the constellation information includes star sign, magnitude, right ascension, and declination.

[0011] The present invention also provides a positioning and navigation method of a starlight navigation device, which is applied to the above-mentioned starlight navigation device. The positioning and navigation method of the starlight navigation device comprises:

[0012] S1, constructing a navigation star catalog by using the star catalog memory;

[0013] S2, taking starry sky picture information through the lens, and sending the starry sky picture information to the computing circuit;

[0014] S3, the calculation circuit calculates and obtains a feature vector of the starry sky picture information according to the starry sky picture information;

[0015] S4. The star catalog memory receives the feature vector, matches the constellation information corresponding to the feature vector, returns the constellation information to the calculation circuit, and the calculation circuit sends the constellation information to the aircraft to calculate the position information.

[0016] In one embodiment of the present invention, constructing a navigation star catalog by using the star catalog memory in step S1 includes:

[0017] S11, selecting stars whose visible stars are smaller than a set threshold according to the characteristics of the imaging original;

[0018] S12, deleting stars whose magnitude changes exceed a set threshold from the selected stars, and if the angle between any two stars is less than the set threshold, both stars are deleted to obtain a navigation star catalog;

[0019] S13, storing the navigation star catalog in a star catalog memory;

[0020] S14. Construct a graph structure relationship according to the relative position information of the navigation star catalog, and perform feature mapping of graph nodes according to the graph structure relationship.

[0021] In one embodiment of the present invention, the calculation circuit in step S3 calculates the feature vector of the starry sky picture information according to the starry sky picture information, including:

[0022] S31, the computing circuit filters the starry sky picture information to obtain filtered starry sky picture information; selects multiple node sets with confidence levels higher than a set threshold in the graph structure of the filtered starry sky picture information for query, and uses the node corresponding to the starry sky picture information with the highest number of hits as the output starry sky picture information according to the query result of the node set;

[0023] S32, constructing a graph structure of the output starry sky picture information;

[0024] S33. Use the GCN model to extract features of the nodes in the graph structure to obtain a feature vector of the starry sky picture information.

[0025] In one embodiment of the present invention, the star catalog memory in step S4 receives the feature vector, matches the constellation information corresponding to the feature vector, returns the constellation information to the calculation circuit, and the calculation circuit sends the constellation information to the aircraft. The calculation position information includes:

[0026] S41, searching and locating the graph structure according to the feature vector to obtain positioning coordinates;

[0027] S42: The computing circuit sends the positioning coordinates to the aircraft to calculate position information.

[0028] In one embodiment of the present invention, the step S41 of retrieving and locating the graph structure according to the feature vector to obtain the location coordinates includes:

[0029] S411, establishing a search association between the starry sky picture information and the node with a fixed numerical value;

[0030] S412, calculating the mapping relationship between the target node and the starry sky picture information by using the jaccard coefficient, and obtaining the candidate position of the target node in the starry sky picture information;

[0031] S413: Implement mutual access between nodes in the starry sky picture information.

[0032] In one embodiment of the present invention, in step S42, the calculation circuit sends the positioning coordinates to the aircraft, and the calculation of the position information includes:

[0033] In the coordinate system of the aircraft, the starlight vector r s By transforming the matrix from the Earth's inertial system to the aircraft coordinate system accomplish: Among them, r s is the starlight vector, is the coordinate system transformation matrix, r i is the position of the star in the Earth’s inertial coordinate system;

[0034] According to the starlight vector r s , calculate the position information of the aircraft.

[0035] The present invention also provides an electronic device, including a processor and a memory, wherein the memory stores program instructions, and the processor executes the program instructions to implement the positioning and navigation method of the starlight navigation device.

[0036] As described above, a starlight navigation device and a positioning and navigation method thereof of the present invention have the following beneficial effects:

[0037] (1) The starlight navigation device of the present invention comprises a sunshade, a lens, a computing circuit, and a star catalog memory. The present invention can calculate the angle of the aircraft relative to the observed star based on the starlight vector measured by the starlight navigation device, thereby calculating the attitude angle of the aircraft and realizing starlight navigation.

[0038] (2) The positioning and navigation method of the starlight navigation device of the present invention can provide position services for objects moving at ultra-high speeds, will not be affected by the external light environment, and the calculated attitude angle of the aircraft is relatively accurate.

[0039] (3) The positioning and navigation method of the starlight navigation device of the present invention constructs a graph structure relationship for the star catalog in the navigation library according to its relative position information, performs feature mapping of the graph nodes according to the generated graph structure, and accelerates the overall retrieval speed with the help of the numerical relationship of the feature mapping. When processing the captured starry sky picture, since a lot of unmatched information may appear due to the noise information that may be contained inside, in order to improve the accuracy of its matching, it is necessary to perform a secondary screening in the original image shooting results. During the screening, the obtained high confidence points are used as references for multi-step searches, and the final positioning results are calculated based on the hit status of the nodes. The current coordinates are calculated through the positioning results to achieve the effect of precise positioning and navigation. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 A schematic diagram of the structure of a starlight navigation device provided in an embodiment of the present invention.

[0041] Figure 2 This is a workflow diagram of a positioning and navigation method for a starlight navigation device provided in an embodiment of the present application.

[0042] Figure 3 A diagram showing the working principle of the positioning and navigation method of the starlight navigation device provided in an embodiment of the present application.

[0043] Component number description

[0044] 1 Lens hood

[0045] 2 Lenses

[0046] 3 Calculation Circuit

[0047] 4 Star catalog memory DETAILED DESCRIPTION

[0048] The following describes the embodiments of the present invention by specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the following embodiments and features in the embodiments can be combined with each other without conflict.

[0049] It should be noted that the illustrations provided in the following embodiments are only used to illustrate the basic concept of the present invention in a schematic manner, and thus the illustrations only show components related to the present invention rather than being drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component may be changed arbitrarily, and the component layout may also be more complicated.

[0050] See also Figure 1 , Figure 1 A schematic diagram of the structure of a starlight navigation device provided by an embodiment of the present invention. The present invention provides a starlight navigation device, comprising a sunshade 1, a lens 2, a computing circuit 3, and a star catalog memory 4, wherein the sunshade 1 is used to filter light pollution generated by an aircraft; the lens 2 is used to obtain starry sky picture information, the sunshade 1 is connected to the lens 2, and the sunshade 1 is arranged at the front end of the lens 2; the computing circuit 3 is used to receive the starry sky picture information and calculate the feature vector of the starry sky picture information; the star catalog memory 4 is used to receive the feature vector and match the constellation information corresponding to the feature vector, and return the constellation information to the computing circuit 3, which sends the constellation information to the aircraft to calculate the position information.

[0051] Specifically, the starlight navigation device of the present invention reduces the light pollution generated by the aircraft itself through the sunshade 1, takes a picture of the starry sky by the lens 2, obtains the starry sky picture information, and transmits the obtained starry sky picture information to the calculation circuit 3 to calculate the characteristic vector of the starry sky picture information, and performs a query and match to the star catalog memory 4. After the matching is completed, the constellation information corresponding to the characteristic vector is returned to the calculation circuit 3, and the constellation information includes the star sign, star magnitude, right ascension, and declination. Through the constellation information, according to the starlight vector measured by the starlight navigation device, the angle of the aircraft relative to the observed star can be calculated, thereby calculating the attitude angle of the aircraft to achieve starlight navigation.

[0052] A graph structure relationship is constructed for the navigation library, i.e., the star table in the star table memory 4, according to its relative position information, and feature mapping of the graph nodes is performed according to the generated graph structure relationship, and the overall retrieval speed is accelerated by means of the numerical relationship of the feature mapping.

[0053] When processing the captured starry sky pictures, there may be a lot of unmatched information due to the noise information contained inside. In order to improve the accuracy of the matching, it is necessary to perform a secondary screening in the original image shooting results. During the screening, the obtained high confidence points are used as references for multi-step search, and the final positioning results are calculated based on the hit status of the nodes. The current coordinates are calculated through the positioning results to achieve the effect of positioning navigation.

[0054] The starlight navigation device of the present invention comprises a sunshade, a lens, a computing circuit, and a star catalog memory. The present invention can calculate the angle of the aircraft relative to the observed star based on the starlight vector measured by the starlight navigation device, thereby calculating the attitude angle of the aircraft to achieve starlight navigation.

[0055] See also Figure 2 , Figure 3 , Figure 2 This is a workflow diagram of a positioning and navigation method for a starlight navigation device provided in an embodiment of the present application. Figure 3 A working principle diagram of a positioning and navigation method of a starlight navigation device provided in an embodiment of the present application. Similar to the principle of a starlight navigation device of the present invention, an embodiment of the present invention further provides a positioning and navigation method of a starlight navigation device, which is applied to the above-mentioned starlight navigation device, and the positioning and navigation method of the starlight navigation device includes:

[0056] Step S1, constructing a navigation star catalog through the star catalog memory 4.

[0057] Specifically, establishing a star catalog memory 4 is a prerequisite for starry sky picture recognition, and it is necessary to take into account star selection principles, storage capacity, storage methods and other conditions. When the number of navigation stars is not enough or the selection is unreasonable, the stars photographed in a certain field of view will not be navigation stars or the number of navigation stars will be less than 3, which will eventually lead to matching failure; and when the number of selected navigation stars is too large, on the one hand, it will lead to an increase in storage capacity and a longer recognition time, and on the other hand, it will also cause the problem of redundant matching.

[0058] The step S1 of constructing a navigation star catalog by means of the star catalog memory 4 includes:

[0059] Step S11, selecting stars with visible stars less than a set threshold value according to the characteristics of the imaging original. Specifically, it can be but not limited to selecting stars with visible stars less than 50 according to the characteristics of the imaging original in a computer.

[0060] Step S12: Delete the stars whose magnitude changes exceed the set threshold value from the selected stars, and if the angle between any two stars is less than the set threshold value, both stars are deleted to obtain a navigation star catalog. Specifically, the stars with large magnitude changes are filtered out, for example, if the angle between two stars is less than 30 arc seconds, both stars are not selected.

[0061] Step S13, storing the navigation star catalog in the star catalog memory 4.

[0062] Step S14: construct a graph structure relationship according to the relative position information of the navigation star catalog, and perform feature mapping of graph nodes according to the graph structure relationship.

[0063] A graph structure relationship is constructed for the star table in the star table storage 4 according to its relative position information, feature mapping of the graph nodes is performed according to the generated graph structure, and the overall retrieval speed is accelerated by means of the numerical relationship of the feature mapping.

[0064] Step S2, photographing starry sky picture information through the lens 2, and sending the starry sky picture information to the calculation circuit 3.

[0065] Step S3: The calculation circuit 3 calculates a feature vector of the starry sky picture information according to the starry sky picture information.

[0066] The calculation circuit 3 in step S3 calculates the feature vector of the starry sky picture information according to the starry sky picture information, including:

[0067] S31, the computing circuit 3 filters the starry sky picture information to obtain filtered starry sky picture information; selects multiple node sets with confidence levels higher than a set threshold in the graph structure of the filtered starry sky picture information for query, and according to the query results of the node sets, uses the node corresponding to the starry sky picture information with the highest number of hits as the output starry sky picture information.

[0068] Specifically, a starry sky picture is taken through lens 2, and the starry sky picture information is transmitted to the calculation circuit 3, and the calculation circuit 3 performs a preliminary screening of the starry sky information in the starry sky picture information. In the algorithm calculation, it is necessary to take the observed star as the main star, and the remaining stars as companion stars. The corresponding navigation star pair is found within a given threshold for the angular distance between the main star and the companion star, and then the brightness level is matched, the matching group is retained, and the next star is selected as the auxiliary star. Deleting the redundant matching star group is to delete the star group that does not meet the threshold value or the brightness does not match. Specifically, magnitude is a measure of the brightness of celestial bodies. In order to measure the brightness of a star, the smaller the magnitude value, the brighter the star; the larger the magnitude value, the darker its light. In the absence of clear instructions, the magnitude generally refers to the visual magnitude. In the present application, it is possible but not limited to screening the star image with a measured magnitude less than 6.5mag on the image of the starry sky picture information as the calibration star to be matched.

[0069] Specifically, when extracting star features, it is necessary to solve the interference of abnormal noise, such as the appearance of pseudo bright spots or the invisible situation of a certain star. It is necessary to include noise data in the sub-graph, or complete high-speed retrieval when the sub-graph is not completely consistent with the query graph. The sub-graph refers to the graph structure generated by the current image, and the query graph refers to the graph structure in the star catalog memory 4, and sub-graph matching is performed. Since the reference information for star feature extraction is only the currently acquired image information, it is impossible to perform accurate noise screening through the information input acquired by itself, so the feature extraction process needs to be post-placed. In the early stage of feature extraction, multiple point sets with high confidence in the graph structure are selected for query, and the starry sky image points with the highest number of hits are used as the final output according to the query results of the point set.

[0070] S32, constructing a graph structure for outputting the starry sky picture information.

[0071] Specifically, in the feature expression of the starry sky image, edges are added to the K nodes with the closest positional relationship in the star table, and a graph structure described by relative positional relationship is constructed through the relationship between nodes. In this way, the robustness of positional relationships such as scale and angle is ensured. When constructing the star table graph structure, the relative positional relationship is used as a reference between nodes. For nodes that are close in distance, the destination can be reached with fewer hops. The introduction of this multi-level node can more completely describe the distance and angle association of nodes.

[0072] Since the current reference information only has the currently acquired image information, it is impossible to accurately screen the noise through the existing information input. Therefore, the currently extracted nodes have the characteristics of disorder, weak dependence and transformation invariance: the current node is a set of points without a specific order, that is, N point sets will have N! ! kinds of arrangements; the nodes come from a space with distance measurement, which means that the points are not independent, and the adjacent points of the current node may form a meaningful subset. The point set is rigidly changed, and the relative position relationship between the points remains unchanged. Rigid changes refer to that only the position (translation transformation) and orientation (rotation transformation) of the object change, while the shape remains unchanged. The resulting transformation is called a rigid transformation.

[0073] We need to use the GCN model to accurately screen the noise of nodes, capture the local structure and the characteristics of the local structures, and achieve accurate feature extraction.

[0074] S33. Use the GCN model to extract features of the nodes in the graph structure to obtain a feature vector of the starry sky picture information.

[0075] Specifically, the GCN model is used to extract node features in the graph structure, and the GCN model is used to perform confidence evaluation for each input node to obtain a feature vector of the starry sky picture information.

[0076] The input of the GCN model is the obtained graph structure. The GCN model is used to evaluate the confidence of each input node and extract the feature vector. The nodes are added with edges based on the pixel distance relationship of the N closest stars to each other, and the three-level connected subgraph structure (CRCN) is extracted from the high-confidence points in turn. The feature vectors are spliced ​​into the final query vector according to the correlation sorting between the nodes for node screening.

[0077] Specifically, a star with a confidence of more than 0.7 is randomly selected as the starting node for random walk as the input of the training sequence, and finally the feature mapping (Embedding) of all nodes is completed. In order to ensure that the feature mapping can express the homogeneity between nodes, the node walk process needs to be biased towards breadth-first search, which can ensure that the nodes in the sequence are most likely to have direct associations. Therefore, the sample information can be more focused on the feature extraction of attribute approximation in feature sampling, while the structurality is reflected in the depth-first search. The depth-first search tends to move to the remote nodes with indirect associations. The long-distance association can make the sample sequence contain more network structure information. According to the obtained confidence evaluation, the nodes are added edges to the nearest N stars according to the pixel distance relationship, that is, the nearest N stars to each other, and the three-level connected subgraph structure (CRCN) is extracted from the high-confidence points in turn. The feature vectors are spliced ​​into the final query vector according to the correlation sorting between the nodes for node screening.

[0078] Step S4, the star catalog memory 4 receives the feature vector, matches the constellation information corresponding to the feature vector, returns the constellation information to the calculation circuit 3, and the calculation circuit 3 sends the constellation information to the aircraft to calculate the position information.

[0079] Specifically, the star catalog memory 4 in step S4 receives the feature vector, matches the constellation information corresponding to the feature vector, returns the constellation information to the calculation circuit 3, and the calculation circuit 3 sends the constellation information to the aircraft. The calculated position information includes:

[0080] Step S41: retrieve and locate the graph structure according to the feature vector to obtain the location coordinates.

[0081] Step S42: The calculation circuit 3 sends the positioning coordinates to the aircraft to calculate the position information.

[0082] In step S41, the graph structure is retrieved and located according to the feature vector to obtain the location coordinates, including:

[0083] Step S411: Establish a search association between starry sky picture information and nodes with a fixed numerical value.

[0084] Step S412: Calculate the mapping relationship between the target node and the starry sky picture information using the Jaccard coefficient to obtain a candidate position of the target node in the starry sky picture information.

[0085] Step S413: realizing mutual access between nodes in the starry sky picture information.

[0086] Specifically, step S411 may include feature mapping quantization: in order to reduce the additional computational error introduced by the similarity measurement between feature maps in the matching process, a retrieval association between the starry sky image and the star points is established with reference to (CGNN changes the numerical relationship to an integer numerical mapping) a fixed numerical size, such as mapping the feature to an N-bit binary or octal code, and completing the rapid retrieval of candidate points through the code.

[0087] Specifically, step S412 may include feature mapping quantization: candidate position hit: during retrieval, the mapping relationship between the target point and the starry sky picture is calculated using the jaccard coefficient (by hit relationship of the base digit) and sorted to obtain the candidate position of the target point in the starry sky picture.

[0088] Specifically, step S413 may include result voting: since the target points are close to each other, mutual access between points can be completed through 2 to 3 jump operations in the starry sky picture, so k steps are walked in the starry sky picture (graph structure), and the point with the highest number of hits is used as the final positioning point.

[0089] The above steps S411 to S413 are all implemented in a computer.

[0090] This process achieves the requirements of real-time and feature uniqueness. Real-time requirement: The matching between the starry sky image and the main star point needs to be completed in a short time. The calculation content includes the extraction and matching of star features, and two methods, global positioning and limited range, can be specified in the matching process. Feature uniqueness: The starry sky image can dynamically limit the search range, and the feature representation of the starry sky image will not change due to the limited range.

[0091] In step S42, the calculation circuit 3 sends the positioning coordinates to the aircraft, and the calculation of the position information includes:

[0092] In the coordinate system of the aircraft, the starlight vector r s By transforming the matrix from the Earth's inertial system to the aircraft coordinate system accomplish: Among them, r s is the starlight vector, is the coordinate system transformation matrix, r i is the position of the star in the Earth’s inertial coordinate system;

[0093] The starlight vector r measured according to the vector s , the angle of the aircraft relative to the observed star can be calculated, and thus the attitude angle of the aircraft can be calculated to achieve starlight navigation.

[0094] In summary, the positioning and navigation method of the starlight navigation device of the present invention constructs a graph structure relationship for the star catalog in the navigation library according to its relative position information, performs feature mapping of the graph nodes according to the generated graph structure, and accelerates the overall retrieval speed with the help of the numerical relationship of the feature mapping. When processing the captured starry sky picture, since a lot of unmatched information may appear due to the noise information that may be contained inside, in order to improve the accuracy of its matching, it is necessary to perform a secondary screening in the original image shooting results. During the screening, the obtained high confidence points are used as references for multi-step searches, and the final positioning results are calculated based on the hit status of the nodes. The current coordinates are calculated through the positioning results to achieve the effect of precise positioning and navigation.

[0095] The above embodiments are merely illustrative of the principles and effects of the present invention, and are not intended to limit the present invention. Anyone familiar with the art may modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by a person of ordinary skill in the art without departing from the spirit and technical concept disclosed by the present invention shall still be covered by the claims of the present invention.

Claims

1. A starlight navigation device, characterized in that: include: A sunshade (1) for filtering light pollution generated by the aircraft; A lens (2) for acquiring starry sky picture information, the light shield (1) being connected to the lens (2), and the light shield (1) being arranged at the front end of the lens (2); A calculation circuit (3), used for receiving the starry sky picture information and calculating a feature vector of the starry sky picture information; The star catalog memory (4) is used to receive the feature vector, match the constellation information corresponding to the feature vector, and return the constellation information to the calculation circuit (3). The calculation circuit (3) sends the constellation information to the aircraft to calculate the position information.

2. A starlight navigation device according to claim 1, characterized in that: The constellation information includes star sign, magnitude, right ascension, and declination.

3. A positioning and navigation method for a starlight navigation device, characterized in that: Applicable to the starlight navigation device according to any one of claims 1 or 2, the positioning and navigation method of the starlight navigation device comprises: S1, constructing a navigation star catalog by using the star catalog memory (4); S2, photographing starry sky image information through the lens (2), and sending the starry sky image information to the calculation circuit (3); S3, the calculation circuit (3) calculates and obtains a feature vector of the starry sky image information according to the starry sky image information; S4. The star catalog memory (4) receives the feature vector, matches the constellation information corresponding to the feature vector, returns the constellation information to the calculation circuit (3), and the calculation circuit (3) sends the constellation information to the aircraft to calculate the position information.

4. The positioning and navigation method of a starlight navigation device according to claim 3, characterized in that: The step S1 of constructing a navigation star catalog by means of the star catalog memory (4) comprises: S11, selecting stars whose visible stars are smaller than a set threshold according to the characteristics of the imaging original; S12, deleting stars whose magnitude changes exceed a set threshold from the selected stars, and if the angle between any two stars is less than the set threshold, both stars are deleted to obtain a navigation star catalog; S13, storing the navigation star catalog in a star catalog memory (4); S14. Construct a graph structure relationship according to the relative position information of the navigation star catalog, and perform feature mapping of graph nodes according to the graph structure relationship.

5. The positioning and navigation method of a starlight navigation device according to claim 3, characterized in that: The calculation circuit (3) in step S3 calculates the feature vector of the starry sky image information according to the starry sky image information, including: S31, the computing circuit (3) filters the starry sky picture information to obtain filtered starry sky picture information; selects a plurality of node sets with confidence levels higher than a set threshold in the graph structure of the filtered starry sky picture information for query, and selects the node corresponding to the starry sky picture information with the highest number of hits as the output starry sky picture information according to the query result of the node set; S32, constructing a graph structure of the output starry sky picture information; S33. Use the GCN model to extract features of the nodes in the graph structure to obtain a feature vector of the starry sky picture information.

6. The positioning and navigation method of a starlight navigation device according to claim 3, characterized in that: The star catalog memory (4) in step S4 receives the feature vector, matches the constellation information corresponding to the feature vector, returns the constellation information to the calculation circuit (3), and the calculation circuit (3) sends the constellation information to the aircraft. The calculated position information includes: S41, searching and locating the graph structure according to the feature vector to obtain positioning coordinates; S42, the computing circuit (3) sends the positioning coordinates to the aircraft to calculate the position information.

7. A positioning and navigation method for a starlight navigation device according to claim 6, characterized in that: In step S41, the graph structure is retrieved and located according to the feature vector to obtain the location coordinates, including: S411, establishing a search association between the starry sky picture information and the node with a fixed numerical value; S412, calculating the mapping relationship between the target node and the starry sky picture information by using the jaccard coefficient, and obtaining the candidate position of the target node in the starry sky picture information; S413: Implement mutual access between nodes in the starry sky picture information.

8. The positioning and navigation method of a starlight navigation device according to claim 6, characterized in that: In step S42, the calculation circuit (3) sends the positioning coordinates to the aircraft, and the calculated position information includes: In the coordinate system of the aircraft, the starlight vector r s By transforming the matrix from the Earth's inertial system to the aircraft coordinate system accomplish: Among them, r s is the starlight vector, is the coordinate system transformation matrix, r i is the position of the star in the Earth’s inertial coordinate system; According to the starlight vector r s , calculate the position information of the aircraft.

9. An electronic device, comprising a processor and a memory, wherein the memory stores program instructions, characterized in that: The processor runs program instructions to implement the positioning and navigation method of the starlight navigation device as described in any one of claims 1 to 7.

Images