Constellation fault source detection method based on graph theory

Abstract

The invention relates to a constellation fault source detection method based on a graph theory, which combines the characteristics of a constellation topological structure, finds the correlation of constellation topological correlation, sets limited monitoring nodes in a constellation to solve the fault probability of satellites in a whole network, and realizes the detection of a fault source satellite. According to the method, the health state of the whole constellation is monitored and detected through measurement and control of limited nodes in the constellation by utilizing the propagation characteristics of inter-satellite links, the dependence on a ground measurement and control station is reduced, and the burden of the ground measurement and control station is relieved; fault source diagnosis is carried out through the thought of the shortest path, the given algorithm can rapidly achieve reachable path weight calculation between nodes, high in-orbit application conditions are achieved, and the autonomy of constellations is improved.

Description

Technical Field

[0001] The present invention belongs to the field of constellation fault diagnosis, and relates to a constellation fault source detection method based on graph theory. Background Art

[0002] In recent years, low-earth-orbit (LEO) mega-constellation systems represented by OneWeb and Starlink have received extensive attention. Compared with traditional medium- and high-earth-orbit satellite constellations, LEO mega-constellations have a low orbital altitude and a large number of satellites. They have the advantages of a wide coverage range, sufficient bandwidth, and relatively low latency. Therefore, LEO mega-constellations have become a very promising satellite constellation construction model. At the same time, the large number of satellites and low manufacturing cost limit the satellite lifespan. As the initial satellite lifespan approaches, the probability of satellite failure becomes higher and higher. The member satellites of LEO mega-constellations are globally network-distributed, with inter-satellite link communication measurement capabilities, and some satellites also rely on the positions of surrounding satellites for positioning and navigation. Therefore, the positions and attitudes of satellites in the constellation are affected by other surrounding satellites, which leads to the "propagation" of faults within the constellation. If the position of the damaged satellite in the constellation cannot be judged in time, it will cause the deviation of local satellite groups or even the entire constellation, and even lead to catastrophic consequences such as chain collisions. Therefore, in order to ensure the safety requirements of LEO mega-constellations in orbit, establishing a constellation-level fault propagation model and effectively identifying the source satellite of the fault in a timely manner are important issues that need to be solved urgently. In the aerospace field, there is currently only fault prediction and health management for single satellites, and there is no relevant research on constellation-level fault diagnosis and fault source identification. The fault detection of single satellites has no correlation with the constellation fault source detection method. Summary of the Invention

[0003] The technical problem solved by the present invention is: overcoming the deficiencies of the prior art, and proposing a constellation fault source detection method based on graph theory. Combining the characteristics of the constellation topology structure, by finding the correlation of the constellation topology association, setting limited monitoring nodes within the constellation to calculate the fault probabilities of all satellites in the network and realizing the detection of the fault source satellite.

[0004] The technical solution adopted by the present invention is:

[0005] In the first aspect, the present invention provides a constellation fault source detection method based on graph theory, including:

[0006] According to past records and combined with the constellation configuration, obtain the fault probability U of each satellite i , and the fault transfer matrix P = {P ij}, i, j = 1, 2,... n, where n is the total number of satellites in the constellation, and P ij represents the fault transfer coefficient between satellite X i and X j in the constellation;

[0007] Determine the topological structure diagram of the directed association network according to the constellation configuration diagram;

[0008] Install several monitoring satellites in the constellation. Starting from the monitoring satellite nodes, according to the satellite reachable path weight solving algorithm, enumerate the reachable paths of each satellite in the constellation to reach the monitoring satellite nodes with the minimum path weight, and calculate the fault transfer probability of each reachable path;

[0009] Compare the path transfer probabilities of each satellite to the monitoring satellite nodes with a logical value of "0" and the path transfer probabilities to the monitoring satellite nodes with a logical value of "1", and determine the fault source sequence for the satellite that propagates to the monitoring satellite nodes with a logical value of "0" with the maximum probability and to the monitoring satellite nodes with a logical value of "1" with the minimum probability.

[0010] Preferably, in the topological structure diagram of the directed association network, for any two connected nodes X i , X j the connection weight is C ij , C ij ≥0, and C ii = 1 (i,j = 1,2,…,n);

[0011] If X i and X j are unidirectionally transmitted, and the transmission direction is X j → X i , or X i and X j are not directly connected, then record C ij = ∞.

[0012] Preferably, according to the satellite reachable path weight solving algorithm, solve the reachable path of satellite X i in the constellation to reach the monitoring satellite X e with the minimum path weight, and the method is as follows:

[0013] S3.1: Determine the connection weights between each satellite node according to the topological structure diagram of the directed association network, and all connection weights form a matrix C; if the logical value of the monitoring satellite X e is "0", then take the e-th column in the matrix C as the initial column vector, and define the initial function f1(i) = C ie , i = 1,2,…,n; f1(i) is the connection weight between X i and X e ;

[0014] S3.2: Use the formula to perform function space iterative operations, k = 2,3,…; where f k-1 (j) is from the monitoring satellite node X eThe minimum value of the product of path weights among all k - 1 - step reachable paths to satellite node j, C ji is the connection weight between satellite node j and satellite node i; f k (i) is the minimum value of the product of path weights among all k - step reachable paths from the monitoring satellite node X e to satellite node i;

[0015] Until f k+1 (i) = f k (i), i = 1, 2, …, n, then f k (i) is the minimum value of the product of path weights among all reachable paths from satellite X i to the monitoring satellite X e ; the reachable path corresponding to f k (i) is the reachable path for satellite X i to reach the monitoring satellite X with the minimum path weight e .

[0016] Preferably, the constellation is composed of multiple satellites operating in low - earth - orbit circular orbits.

[0017] Preferably, the reachable path is a non - closed - loop path, that is, starting from the source node, moving forward in the direction of the arrows in the directed associated network topology structure diagram to reach the target node, and any node is passed through at most once on the way.

[0018] Preferably, the satellite status of the monitoring satellite is represented by a logical value, and the logical value is 1 or 0. When the logical value of the monitoring satellite is 0, it indicates that there is a satellite failure in the constellation.

[0019] Preferably, the logical value of the monitoring satellite can be obtained by ground telemetry and control.

[0020] In a second aspect, the present invention provides a terminal device, including:

[0021] A memory for storing instructions executed by at least one processor;

[0022] A processor for executing the instructions stored in the memory to implement the method described in the first aspect above.

[0023] In a third aspect, the present invention provides a computer - readable storage medium, and the computer - readable storage medium stores computer instructions. When the computer instructions run on a computer, the computer is caused to execute the method described in the first aspect above.

[0024] The beneficial effects of the present invention compared with the prior art are:

[0025] The present invention provides a constellation fault source detection method based on graph theory, which establishes a shortest path detection method for constellation-level fault propagation, determines the probability of each satellite reaching a fault node, can effectively identify and isolate the fault source satellite in a timely manner, and improves the safety of the giant constellation in orbit; by utilizing the propagation characteristics of inter-satellite links, the health status monitoring and detection of the entire constellation are realized through the measurement and control of a limited number of nodes in the constellation, reducing the dependence on ground measurement and control stations and alleviating the burden on ground measurement and control stations; the shortest path method is used for fault source diagnosis, and the algorithm given can quickly calculate the weight of the reachable path between nodes, has high on-orbit application conditions, and improves the autonomy of the constellation. Description of the Drawings

[0026] Figure 1 is the configuration of a certain constellation;

[0027] Figure 2 is Figure 1 the topological structure diagram of the directed incidence network of. Detailed Embodiment

[0028] The present invention will be further described below with reference to the drawings.

[0029] In the present invention, the constellation is composed of multiple satellites operating in low-earth circular orbits, with inter-satellite link communication measurement capabilities, and some satellites also rely on the positions of surrounding satellites for positioning and navigation. Therefore, the positions and attitudes of satellites in the constellation are affected by other surrounding satellites. If a satellite node fails, the fault will propagate along the inter-satellite link. The reachable path of constellation fault propagation is defined as follows: The reachable path is a non-closed path, that is, starting from the source node, moving forward in the direction of the network arrow to reach the target node, and any node is passed through at most once on the way.

[0030] The constellation fault source detection task can be described as: a satellite constellation N = {X1, X 2… X n} composed of n node satellites, and the fault transfer matrix P = {P 2… X n} ∈ R ij , where P n×n represents the fault transfer coefficient between satellite X ij and X i . R j represents an n×n real matrix. n×n

[0031] There are several monitoring satellites in the constellation, and their satellite states (1 or 0) can be obtained by ground TT&C. When there is a 0 in the monitoring satellite state, it means that there is a satellite failure in the constellation. Then, it is necessary to find the satellite node of the fault source in constellation N. The fault of this node can be propagated to the monitoring satellite node with a logical value of "0" with the highest probability, and at the same time, it is propagated to the monitoring satellite node with a logical value of "1" with the lowest probability.

[0032] The present invention provides a terminal device, including: a memory for storing instructions executed by at least one processor; a processor for executing the instructions stored in the memory to implement the above method.

[0033] The present invention also provides a computer-readable storage medium storing computer instructions, which when run on a computer, cause the computer to execute the above method.

[0034] Embodiment:

[0035] Taking Figure 1 the shown constellation configuration as an example for illustration, the steps for determining the fault source are as follows:

[0036] S1: According to past records and experience, combined with the constellation configuration, obtain the fault probability U of each satellite i (i = 1, 2,... n), and the inter-satellite fault transfer matrix P = {P ij} ∈ R n×n (i, j = 1, 2,... n).

[0037] According to the constellation configuration diagram, the method for calculating the inter-satellite fault transfer matrix is as follows:

[0038] Considering that the inter-satellite link transfer of co-orbit satellites is bidirectional, and the inter-satellite link transfer between non-co-orbit satellites is unidirectional, and the configurations of the inter-satellite links are all the same, that is, the propagation probability of each link is the same. It can be obtained that Figure 1 the inter-satellite transfer matrix of directly connected satellites in the constellation is shown in Table 1.

[0039] Table 1 Constellation fault transfer matrix

[0040]

[0041] S2: According to the constellation configuration diagram, draw the system directed association network topology structure diagram, such as Figure 2 , and the weight value in the directed association network topology structure diagram is the reciprocal of the fault transfer coefficient, that is, X i →X j , C ij = 1 / P ij .

[0042] S3: FromFigure 2 , starting from the monitoring satellite nodes (Star 2, Star 8, Star 12, and Star 18 in this case), according to the satellite reachable path weight solution algorithm, enumerate the reachable paths of each satellite in the constellation to reach the monitoring satellite nodes with the minimum path weight, and calculate the path fault transfer probability.

[0043] The minimum weight solution of the fault propagation reachable path can be solved by the space iteration method. This task can be described as follows: A satellite constellation N = {X1, X2…X n} composed of n node satellites, the connection weight between any two nodes X i and X j is C ij , C ij ≥0, and C ii = 1 (i, j = 1, 2,…, n). If X i and X j are not adjacent (no direct connection), or X i and X j have a one-way transfer, and the transfer direction is X j → X i , then record C ij = ∞. The reachable path solution is to find the path with the minimum product of path weights and the minimum weight among the reachable paths from any satellite X i to the target star X e . The solution steps are as follows:

[0044] S3.1: Determine the initial function. According to the constellation network structure, determine the connection weight C ij between each node, as shown in Table 2. If the logical value of the monitoring satellite X e is "0", when finding its reachable path, take the e-th column in the C matrix as the initial function f1(i) = C ei , i = 1, 2,…, n. This column is called the initial column vector. f1(i) is the connection weight between X i and X e .

[0045] Table 2 State connection weight table (C matrix)

[0046]

[0047] S3.2: Use the formula for function space iteration operation, k = 2, 3,…; where f k-1 (j) is the minimum value of the product of path weights among all k - 1 step reachable paths from the monitoring satellite node X e to the satellite node j, C ji is the connection weight between the satellite node j and the satellite node i; f k(i) is the minimum value of the product of path weights among all k-step reachable paths from the monitoring satellite node X e to the satellite node i; among them, it is easy to know that f k (e)=1.

[0048] S3.3: Repeat step S3.2 until f k+1 (i)=f k (i), i = 1, 2, …, n, then f k (i) is the minimum value of the product of path weights among all reachable paths from satellite X i to satellite X e The reachable path corresponding to f k (i) is the reachable path for satellite X i to reach the monitoring satellite X with the minimum path weight e .

[0049] By enumeration, the minimum path weights from each star in the constellation configuration diagram to each monitoring node star (star 2, star 8, star 12, star 18) are

[0050]

[0051] S4: Based on the results of S3, compare the path transfer probabilities of each satellite to the monitoring satellite nodes with a logical value of "0" and the path transfer probabilities to the monitoring satellite nodes with a logical value of "1", and propagate to the monitoring satellite nodes with a logical value of "0" with the maximum probability. At the same time, propagate to the monitoring satellite nodes with a logical value of "1" with the minimum probability to determine the fault source sequence.

[0052] The parts not detailed in the present invention belong to the common general knowledge of those skilled in the art.

Claims

1. A constellation fault source detection method based on graph theory, characterized in that: include: According to previous records and combined with the constellation configuration, the failure probability U of each satellite is obtained. i , and the constellation fault transfer matrix P = {P ij }, i, j = 1, 2, ... n, n is the total number of satellites in the constellation, P ij Indicates satellite X in the constellation i With X j The fault transmission coefficient between According to the constellation configuration diagram, determine the directed association network topology diagram; There are several monitoring satellites in the constellation. Starting from the monitoring satellite node, according to the satellite reachable path weight solution algorithm, the reachable paths of each satellite in the constellation with the minimum path weight to the monitoring satellite node are enumerated, and the fault transmission probability of each reachable path is calculated; The path transmission probability of each satellite reaching the monitoring satellite node with a logical value of "0" is compared with the path transmission probability of each satellite reaching the monitoring satellite node with a logical value of "1". The satellite that propagates to the monitoring satellite node with a logical value of "0" with the maximum probability and propagates to the monitoring satellite node with a logical value of "1" with the minimum probability determines the fault source sequence.

2. The method for detecting constellation fault sources based on graph theory according to claim 1, characterized in that: In the directed association network topology diagram, any two connected nodes X i , X j The connection weight is C ij , And C ii =1(i,j=1,2,…,n); If X i With X j One-way transmission, the transmission direction is X j →X i , or X i With X j If not directly connected, record C ij =∞.

3. The method for detecting constellation fault sources based on graph theory according to claim 1, characterized in that: According to the satellite reachable path weight solution algorithm, solve the satellite X in the constellation i Reach monitoring satellite X with the minimum path weight e The reachable path is as follows: S3.1: Determine the connection weights between satellite nodes according to the directed association network topology diagram, and all connection weights form a matrix C; if the monitoring satellite X e If the logical value is "0", then take the e-th column in the matrix C as the initial column vector and define the initial function f1(i)=C ie ,i=1,2,…,n;f1(i) is X i With X e The connection weight of S3.2: Using formula Perform iterative operation in function space, k = 2, 3, ...; where f k-1 (j) is the monitoring satellite node X e The minimum value of the product of path weights among all k-1-step reachable paths to satellite node j, C ji is the connection weight between satellite node j and satellite node i; f k (i) is from monitoring satellite node X e The minimum value of the product of path weights among all k-step reachable paths to satellite node i; Until f k+1 (i) = f k (i), i=1,2,…,n, then f k (i) is satellite X i To surveillance satellite X e The minimum value of the product of path weights among all reachable paths, f k (i) The corresponding reachable path is satellite X i Reach monitoring satellite X with the minimum path weight e The reachable path.

4. The method for detecting constellation fault sources based on graph theory according to claim 1, characterized in that: The constellation consists of multiple satellites operating in low circular orbits.

5. The method for detecting constellation fault sources based on graph theory according to claim 1, characterized in that: A reachable path is a non-closed loop path, that is, starting from the source node, moving in the direction of the arrow in the directed associated network topology diagram to reach the target node, and passing through any node on the way at most once.

6. The method for detecting constellation fault sources based on graph theory according to claim 1, characterized in that: The satellite status of the monitoring satellite is represented by a logic value, and the logic value is 1 or 0. When the monitoring satellite logic value is 0, it indicates that a satellite in the constellation has a fault.

7. The method for detecting constellation fault sources based on graph theory according to claim 6, characterized in that: The logic value of the monitoring satellite can be obtained by ground measurement and control.

8. A terminal device, characterized in that: include: a memory for storing instructions executed by at least one processor; A processor, configured to execute instructions stored in a memory to implement a method according to any one of claims 1 to 7.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and when the computer instructions are executed on a computer, the computer is enabled to execute the method according to any one of claims 1 to 7.

Images