A Method for Selecting the Contract Net Master Star Based on the Task Area Guarding Strategy

Abstract

The present invention belongs to the technical field of contract network master satellite selection, and relates to a contract network master satellite selection method based on a mission area on-duty strategy, including the following steps: obtaining the space state data, platform state data, and mission state data of all satellites in the constellation within a certain rolling time domain; an adaptive constellation settlement generation method based on the minimum hop count constraint, according to the set minimum hop count, using a spatial density clustering algorithm to adaptively generate a constellation settlement set; through a settlement master satellite selection method based on the mission area on-duty strategy, comprehensively considering the various performances of the satellites, calculating the master satellite and backup master satellite of each settlement set; the present invention optimizes the master satellite selection strategy, and improves the reliability of distributed mission planning and the adaptability and robustness of the distributed system.

Description

Technical Field

[0001] The present invention belongs to the technical field of collaborative observation planning for distributed satellite systems, and particularly relates to a method for selecting a contract network master satellite based on a mission area duty strategy. Background Art

[0002] With the expansion of the constellation scale, centralized task planning algorithms are gradually difficult to apply to the task scheduling and management of large-scale distributed constellation systems. Compared with centralized methods, distributed methods have become a current research hotspot due to their high reliability, good fault tolerance, reconfigurability, high computational efficiency, etc. Among them, the contract network mechanism is widely applied to the collaborative scheduling problem of distributed satellite systems. As a distributed collaborative method based on multi-agent theory, the contract network mechanism effectively solves the online collaborative scheduling problem of distributed satellites through three processes: tendering, bidding, and bid evaluation.

[0003] In the current distributed satellite task planning methods based on the contract network protocol, most researchers mainly focus on research in processes such as tendering, bidding, and bid evaluation. For example: Intelligent task planning for on-orbit distributed negotiation of satellites for instant response [J]. Chinese Journal of Space Science, 2024, 44(01): 159 - 168, Contract network algorithm for task allocation of distributed remote sensing satellites [J]. Journal of Harbin Engineering University, 2020, 41(07): 1059 - 1065. There is little research on the selection of the master satellite in the contract network protocol. In existing research, the selection of the contract network master satellite mainly relies on random selection planning and routing performance priority strategies. For example: Earth observation satellite task planning and scheduling technology [M]. Beijing: National Defense Industry Press, 2021: 112 - 120, A new virtual satellite formation scheme for deterministic transmission [J]. Journal on Communications, 2023, 44(10): 137 - 148. However, the master satellite selection process should comprehensively consider its computing and communication capabilities, orbital coverage, system reliability, and dynamic adjustment capabilities. Random or single selection strategies will lead to unreasonable master satellite selection, thus having an adverse impact on distributed task planning. By optimizing the master satellite selection strategy, it helps to improve the efficiency, fairness of distributed task planning, and the adaptability and robustness of distributed systems. Summary of the Invention

[0004] The purpose of the present invention is to overcome the deficiencies in the prior art and provide a method for selecting a contract network master satellite based on a mission area duty strategy.

[0005] In order to achieve the purpose of the present invention, the following technical solutions are adopted for implementation.

[0006] A method for selecting a contract network master satellite based on a mission area duty strategy includes the following steps:

[0007] S1. Obtain the space state data, platform state data, and mission state data of m satellites from a distributed satellite constellation within a certain rolling time domain window. Among them:

[0008] The space state data is the ECEF space position and velocity of satellite i. Where are the x, y, and z axis positions of satellite i in the ECEF coordinate system respectively, are the x, y, and z axis velocities of satellite i in the ECEF coordinate system respectively.

[0009] The platform state data is the satellite health of satellite i , remaining battery power , remaining memory , remaining bandwidth , denoted as ;

[0010] The mission state data is the mission completion rate of satellite i ;

[0011] S2. Respectively construct a space state data set , a platform state data set , and a mission state data set according to the space state data, platform state data, and mission state data obtained in step S1. While constructing the space state data set , set the initial number of clusters N , and set the minimum number of hops between satellites within a cluster M； The platform state data set includes a platform number set ;

[0012] S3. Randomly select N coordinate points from the space state data set as the initial cluster set , input the initial cluster set into the spatial density clustering algorithm, and set the number of iterations iter and the iteration accuracy error for iterative calculation to obtain the cluster set ;

[0013] S4. Define the link establishment constraint conditions for each satellite according to the geometric visibility constraint, inter-satellite distance constraint, and relative angular velocity constraint that affect the visibility between satellites. The link establishment constraint conditions are: geometric visibility constraint = inter-satellite distance constraint = relative angular velocity constraint = 1, that is , perform link establishment judgment on each satellite in the cluster, and let i = 1;

[0014] S5. Construct of the satellites undirected connected graph of order , where is the vertex set of the connected graph, is the edge set of the connected graph, represents the edge from vertex v j to vertex v k ;

[0015] S6. According to the link establishment constraint conditions in step S4 and the connectivity matrix defined by the undirected connected graph G in step S5, i by performing connectivity judgment on all the satellites in the settlement s i , construct the connectivity matrix W i , which is the routing table of the settlement s i

[0016] S7. Through the Dijkstra algorithm, repeatedly use the iterative formula:

[0017] ;

[0018] After iterative calculation, obtain the minimum hop count matrix mt i , and the expression of the minimum hop count matrix mt i

[0019] ;

[0020] In the formula, is the minimum hop count from vertex to vertex ;

[0021] According to the link establishment judgment results of each satellite in the settlement S, let i = 1, and iteratively calculate the settlement minimum hop count matrix set ; If the largest element in MT is less than the set minimum hop count M, that is , then output the settlement set S, otherwise N = N + 1, and return to step S3;

[0022] S8. Obtain the spatial state data set , the platform state data set , the task state data set and the clustering set S from steps S2 and S3 respectively, and input the ECEF coordinates of the task area ;

[0023] S9. The settlement s i contains the platform number set , csp i,j is the number of the j-th satellite in the settlement s i , and the settlement s iThe spatial coordinates of the center point are denoted as , satellite csp i,j and the distance dp from the settlement center j . There is a settlement s i Center distance matrix ;

[0024] S10. Divide the motion cycle between satellites into a series of time slices with equal intervals based on the virtual topology strategy to construct satellite network snapshots, and then calculate satellite routing performance metrics, satellite platform performance metrics based on the satellite platform data set, and satellite mission performance metrics based on the satellite mission data set; the satellite routing performance metrics include average satellite routing delay, satellite routing delay difference, average node degree, and node degree jitter; the satellite platform performance metrics include satellite health, remaining power, remaining memory, and remaining bandwidth; the satellite mission performance metric is the task backlog rate; where:

[0025] The matrix of the average routing delay is:

[0026] ;

[0027] In the formula, TS is the minimum delay matrix between satellites, is a matrix of all 1s of m*1;

[0028] The matrix of the routing delay difference is:

[0029] ;

[0030] The matrix of the average node degree is:

[0031] ;

[0032] The matrix of the node degree jitter is:

[0033] ;

[0034] In the formula, represents taking the maximum value of each row of MW, represents taking the minimum value of each row of MW;

[0035] The matrix of the satellite health is:

[0036] ;

[0037] The matrix of the remaining power is:

[0038] ;

[0039] The matrix of the remaining memory is:

[0040] ;

[0041] The matrix of the remaining bandwidth is:

[0042] ;

[0043] The matrix of the task backlog rate is:

[0044] ;

[0045] Wherein, is the total number of tasks of satellite m, is the number of unfinished tasks of satellite m;

[0046] S11. According to the satellite routing performance index, satellite platform performance index and satellite task performance index in step S10, construct an index matrix , for settlement s i , substitute into the entropy weight method, calculate and obtain the main star efficiency matrix i of each satellite in settlement s , sort in descending order, and record the efficiency value ranked first as , and record the corresponding satellite number as , the efficiency value ranked second is , and record the corresponding satellite number as , and select the main star ms and the backup main star bs i and backup main star bs i according to the task area duty strategy; wherein:

[0047] The task area duty strategy is constructed according to the known efficiency matrix and the satellite number matrix ; The task area duty strategy is:

[0048] a. If is located within the task area Tg, is located outside the task area Tg, then the main star ms i = , the backup main star bs i = ;

[0049] b. If is located outside the task area Tg, is located within the task area Tg, and , then the main star ms i = , the backup main star bs i = ;

[0050] c. If is outside the mission area Tg, is inside the mission area Tg, and , then the main satellite ms i = , the backup main satellite bs i = ;

[0051] d. If and are both outside the mission area Tg, then the main satellite ms i = , the backup main satellite bs i = ;

[0052] S12. Let i = 1, and iteratively calculate the main satellite set and the backup main satellite set based on step S11.

[0053] As a preferred embodiment of the present invention, the distributed satellite constellation is composed of satellites with different functional payloads and different orbits; among them: the functional payloads include optical remote sensing payloads, SAR payloads, and electronic reconnaissance payloads; the orbit distribution covers low earth orbit, medium earth orbit, and geosynchronous orbit.

[0054] As a preferred embodiment of the present invention, the mission scheduling of the distributed satellite constellation adopts an event-triggered hybrid scheduling framework to perform mission scheduling on satellite resources in different rolling time domains in stages.

[0055] As a preferred embodiment of the present invention, the geometric visibility constraint requires that there is no direct earth occlusion between satellites, and when any satellite is not in the atmosphere, it means that the satellites are geometrically visible. If the satellites are geometrically visible, then count = 1;

[0056] The inter-satellite distance constraint requires that the spatial distance between satellites is less than or equal to the maximum transmission distance, then the satellites satisfy the inter-satellite distance constraint, count ;

[0057] The relative angular velocity constraint requires that the relative operation between satellites does not exceed the antenna acquisition and tracking capabilities of any satellite, then the satellites satisfy the relative angular velocity constraint, count ran = 1.

[0058] As a preferred embodiment of the present invention, the construction process of the routing table of the settlement s i :

[0059] The settlement s i contains A constellation of satellites is constructed based on graph theory An undirected connected graph of order , where is the vertex set of the connected graph, is the edge set of the connected graph, represents the edge from vertex v j to vertex v k .

[0060] Based on the link - building constraint conditions defined in step S4, the connectivity matrix of the undirected connected graph G is defined as:

[0061] ;

[0062] By performing connectivity judgment on all the satellites in settlement s i , a connectivity matrix W i is constructed, which is the routing table of settlement s i .

[0063] As a preferred embodiment of the present invention, the satellite routing performance index is given by the minimum - delay matrix between satellites; where: the calculation process of the minimum - delay matrix includes the following steps:

[0064] S61. Construct an undirected connected graph of all m satellites in the constellation based on graph theory , and calculate the distance l j,k between satellite j and satellite k in the constellation, then there is an inter - satellite distance matrix ;

[0065] S62. Based on the link - building constraint conditions defined in step S4, construct the connectivity matrix and the delay matrix of the undirected connected graph G, and define as:

[0066] ,

[0067] In the formula, c is the speed of light;

[0068] S63. Through the Dijkstra algorithm, calculate the minimum - delay matrix between satellites, denoted as:

[0069] ;

[0070] In the formula, is the minimum delay from vertex to vertex .

[0071] Advantages of the present invention:

[0072] 1. The present invention proposes an adaptive constellation settlement generation method based on the minimum hop count constraint. According to the set minimum hop count, an adaptive constellation settlement is generated using a spatial density clustering algorithm.

[0073] 2. The present invention proposes a process for selecting the main star of a settlement based on a task area duty strategy. By comprehensively considering the various performances of satellites, the main star and the backup main star are dynamically selected through the task area duty strategy. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] Figure 1 is the contract network main star selection process based on the task area duty strategy in the present invention;

[0075] Figure 2 is the adaptive constellation settlement generation process based on the minimum hop count constraint in the present invention;

[0076] Figure 3 is the process for selecting the main star of a settlement based on the task area duty strategy in the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0077] The present invention will be further described in conjunction with the embodiments and the drawings.

[0078] As Embodiment 1 of the present invention, it is set that the total constellation scale is 72, including 48 low-earth orbit optical remote sensing satellites and 24 low-earth orbit SAR satellites, and the length of the rolling time domain window is set to 20 minutes. As Figures 1 to 3 shown, a method for selecting the main star of a contract network based on a task area duty strategy includes the following steps:

[0079] S1. Obtain the spatial state data, platform state data, and task state data of m satellites from the distributed satellite constellation within a certain rolling time domain window; where:

[0080] The spatial state data is the ECEF spatial position and velocity of satellite i , where are the x, y, and z axis positions of satellite i in the ECEF system respectively, are the x, y, and z axis velocities of satellite i in the ECEF system respectively;

[0081] The platform state data is the satellite health of satellite i , remaining power , remaining memory , remaining bandwidth , denoted as ;

[0082] The task state data is the task completion rate of satellite i ;

[0083] S2. Construct a spatial state data set, a platform state data set, and a task state data set respectively according to the spatial state data, platform state data, and task state data obtained in step S1 , the platform state data set , and the task state data set ; While constructing the spatial state data set , set the initial number of settlements 5 , and set the minimum number of hops between satellites within a settlement 3 ; The platform state data set includes a platform number set ;

[0084] S3. Randomly select 5 coordinate points from the spatial state data set as the initial settlement set , input the initial settlement set into the spatial density clustering algorithm, and set the number of iterations iter = 100 and the iteration accuracy error = 0.001 for iterative calculation to obtain the settlement set ;

[0085] S4. Define the link establishment constraint conditions for each satellite according to the geometric visibility constraint, inter-satellite distance constraint, and relative angular velocity constraint that affect the visibility between satellites. The link establishment constraint conditions are: geometric visibility constraint = inter-satellite distance constraint = relative angular velocity constraint = 1, that is , perform link establishment judgment on each satellite in the settlement, and let i = 1;

[0086] S5. Construct an undirected connected graph of order satellites in the settlement according to graph theory , where is the vertex set of the connected graph, is the edge set of the connected graph, represents the edge from vertex v to vertex v j to vertex v k ;

[0087] S6. According to the link establishment constraint conditions in step S4 and the connection matrix defined by the undirected connected graph G in step S5, by performing connection judgment on all satellites in the settlement s i , construct the connection matrix W i , which is the routing table of the settlement s i ;

[0088] S7. Through the Dijkstra algorithm, repeatedly use the iterative formula:

[0089] ;

[0090] After iterative calculation, the minimum hop count matrix mt is obtained i , and the minimum hop count matrix mt i has the following expression:

[0091] ;

[0092] In the formula, is the minimum hop count from vertex to vertex ;

[0093] According to the link establishment judgment results of each satellite in the settlement S, let i = 1, and perform loop iteration to calculate the settlement minimum hop count matrix set ; If the largest element in MT is less than the set minimum hop count of 3, that is , then output the settlement set S, otherwise N = N + 1, and return to step S3;

[0094] S8. Obtain the spatial state data set , platform state data set , task state data set and the clustering set S from steps S2 and S3 respectively, and input the ECEF coordinates of the task area;

[0095] S9. The settlement s i contains the platform number set , csp i,j is the number of the j-th satellite in the settlement s i . Denote the central spatial coordinates of the settlement s i as , and the distance dp i,j between the satellite csp and the settlement center j . There is a central distance matrix i for the settlement s ;

[0096] S10. Based on the virtual topology strategy, divide the motion cycle between satellites into a series of equally spaced time slices to construct satellite network snapshots, and then calculate satellite routing performance metrics, satellite platform performance metrics based on the satellite platform data set, and satellite task performance metrics based on the satellite task data set; The satellite routing performance metrics include satellite average routing delay, satellite routing delay difference, average node degree, and node degree jitter; The satellite platform performance metrics include satellite health, remaining power, remaining memory, and remaining bandwidth; The satellite task performance metric is the task backlog rate; Among them:

[0097] The matrix of the average routing delay is:

[0098] ;

[0099] Wherein, TS is the minimum delay matrix between satellites, is an all-ones matrix of m*1;

[0100] The matrix of the routing delay difference is:

[0101] ;

[0102] The matrix of the average node degree is:

[0103] ;

[0104] The matrix of the node degree jitter is:

[0105] ;

[0106] Wherein, represents taking the maximum value for each row of MW, represents taking the minimum value for each row of MW;

[0107] The matrix of the satellite health is:

[0108] ;

[0109] The matrix of the remaining power is:

[0110] ;

[0111] The matrix of the remaining memory is:

[0112] ;

[0113] The matrix of the remaining bandwidth is:

[0114] ;

[0115] The matrix of the task backlog rate is:

[0116] ;

[0117] Wherein, is the total number of tasks of satellite m, is the number of unfinished tasks of satellite m;

[0118] S11. According to the satellite routing performance index, satellite platform performance index and satellite task performance index in step S10, construct an index matrix for settlement s i , substitute into the entropy weight method, and calculate to obtain settlement s iThe main satellite effectiveness matrix of each satellite in , sort in descending order, and record the effectiveness value ranked first as , and record the corresponding satellite number as . The effectiveness value ranked second is , and the corresponding satellite number is recorded as . Then, select the main satellite ms and the backup main satellite bs i from i according to the mission area on-duty strategy; where:

[0119] The mission area on-duty strategy is constructed based on the known effectiveness matrix and the satellite number matrix . The mission area on-duty strategy is as follows:

[0120] a. If is located within the mission area Tg and is located outside the mission area Tg, then the main satellite ms i = , and the backup main satellite bs i = ;

[0121] b. If is located outside the mission area Tg, is located within the mission area Tg, and , then the main satellite ms i = , and the backup main satellite bs i = ;

[0122] c. If is located outside the mission area Tg, is located within the mission area Tg, and , then the main satellite ms i = , and the backup main satellite bs i = ;

[0123] d. If and are both located outside the mission area Tg, then the main satellite ms i = , and the backup main satellite bs i = ;

[0124] S12. Let i = 1, and calculate the main satellite set and the backup main satellite set through loop iteration based on step S11.

[0125] As Embodiment 2 of the present invention, the total constellation scale is expanded to 102 satellites, including 48 low-earth orbit optical remote sensing satellites, 2 geostationary optical remote sensing satellites, 24 low-earth orbit SAR satellites, and 28 electronic reconnaissance satellites, and the rolling time domain window length is set to 10 minutes. As shown in Figures 1 to 3 A contract network master satellite selection method based on a mission area duty strategy includes the following steps:

[0126] S1. Obtain the space state data, platform state data, and mission state data of 28 satellites from the distributed satellite constellation within the first rolling time domain window; where:

[0127] The space state data is the ECEF space position and velocity of satellite i, where are the x, y, and z axis positions of satellite i in the ECEF coordinate system respectively, and are the x, y, and z axis velocities of satellite i in the ECEF coordinate system respectively; The platform state data is the satellite health of satellite i,

[0128] remaining battery power, remaining memory, remaining bandwidth, denoted as ; ;

[0129] The mission state data is the mission completion rate of satellite i ;

[0130] S2. Construct a space state data set , a platform state data set , and a mission state data set respectively according to the space state data, platform state data, and mission state data obtained in step S1; while constructing the space state data set , set the initial number of settlements 8 , and set the minimum number of hops between satellites within a settlement 3 ; the platform state data set includes a platform number set ;

[0131] S3. Randomly select 8 coordinate points from the space state data set as the initial settlement set , input the initial settlement set into the spatial density clustering algorithm, and set the number of iterations iter = 100 and the iteration accuracy error = 0.001 for iterative calculation to obtain the settlement set ;

[0132] S4. Define the link establishment constraint conditions for each satellite according to the geometric visibility constraint, inter-satellite distance constraint, and relative angular velocity constraint that affect the visibility between satellites. The link establishment constraint conditions are: geometric visibility constraint = inter-satellite distance constraint = relative angular velocity constraint = 1, that is , perform link establishment judgment on each satellite in the settlement, and let i = 1;

[0133] S5. Construct an undirected connected graph of satellites in the settlement according to graph theory of order , where is the vertex set of the connected graph, is the edge set of the connected graph, represents the edge from vertex v j to vertex v k ;

[0134] S6. According to the link establishment constraint conditions in step S4 and the connection matrix defined according to the undirected connected graph G in step S5, i perform connection judgment on all satellites in the settlement s i to construct the connection matrix W i , which is the routing table of the settlement s

[0135] S7. Through the Dijkstra algorithm, repeatedly use the iterative formula:

[0136] ;

[0137] After iterative calculation, obtain the minimum hop count matrix mt i , and the expression of the minimum hop count matrix mt i is:

[0138] ;

[0139] In the formula, is the minimum hop count from vertex to vertex ;

[0140] According to the link establishment judgment results of each satellite in the settlement S, let i = 1, and loop and iterate to calculate the settlement minimum hop count matrix set ; If the largest element in MT is less than the set minimum hop count of 3, that is , then output the settlement set S, otherwise N = N + 1, and return to step S3;

[0141] S8. Obtain the space state data set and the platform state data set , the task status data set and the clustering set S, and input the ECEF coordinates of the task area ;

[0142] S9. Settlement s i contains the platform number set , csp i,j is the number of the j-th satellite in settlement s i . Denote the spatial coordinates of the center point of settlement s i as , satellite csp i,j and the distance dp from the settlement center j . There is a center distance matrix i for settlement s ;

[0143] S10. Divide the movement period between satellites into a series of equally spaced time slices based on the virtual topology strategy to construct a satellite network snapshot, and then calculate the satellite routing performance metrics, calculate the satellite platform performance metrics based on the satellite platform data set, and calculate the satellite task performance metrics based on the satellite task data set; the satellite routing performance metrics include the average satellite routing delay, satellite routing delay difference, average node degree, and node degree jitter; the satellite platform performance metrics include satellite health, remaining power, remaining memory, and remaining bandwidth; the satellite task performance metric is the task backlog rate; where:

[0144] The matrix of the average routing delay is:

[0145] ;

[0146] In the formula, TS is the minimum delay matrix between satellites, is an all-1 matrix of m*1;

[0147] The matrix of the routing delay difference is:

[0148] ;

[0149] The matrix of the average node degree is:

[0150] ;

[0151] The matrix of the node degree jitter is:

[0152] ;

[0153] In the formula, represents taking the maximum value of each row of MW, represents taking the minimum value of each row of MW;

[0154] The matrix of the satellite health is as follows:

[0155] ;

[0156] The matrix of the remaining power is as follows:

[0157] ;

[0158] The matrix of the remaining memory is as follows:

[0159] ;

[0160] The matrix of the remaining bandwidth is as follows:

[0161] ;

[0162] The matrix of the task backlog rate is as follows:

[0163] ;

[0164] Among them, is the total number of all tasks of satellite m, is the number of unfinished tasks of satellite m;

[0165] S11. According to the satellite routing performance index, satellite platform performance index and satellite task performance index in step S10, construct the index matrix , for settlement s i , substitute into the entropy weight method, calculate and obtain the main star efficiency matrix i of each satellite in settlement s . Sort in descending order, and record the efficiency value ranked first as , and record the corresponding satellite number as . The efficiency value ranked second is , and record the corresponding satellite number as . And select the main star ms and the backup main star bs i from i according to the task area duty strategy; where:

[0166] The task area duty strategy is constructed according to the known efficiency matrix and the satellite number matrix ; The task area duty strategy is:

[0167] a. If is located within the task area Tg and is located outside the task area Tg, then the main star ms i = , the backup primary satellite bs i = ;

[0168] b. If is located outside the mission area Tg, is located inside the mission area Tg, and , then the primary satellite ms i = , the backup primary satellite bs i = ;

[0169] c. If is located outside the mission area Tg, is located inside the mission area Tg, and , then the primary satellite ms i = , the backup primary satellite bs i = ;

[0170] d. If and are both located outside the mission area Tg, then the primary satellite ms i = , the backup primary satellite bs i = ;

[0171] S12. Let i = 1, and iteratively calculate the primary satellite set and the backup primary satellite set .

[0172] As Embodiment 3 of the present invention, the total constellation scale is expanded to 202 satellites, including 96 low-orbit optical remote sensing satellites, 2 geosynchronous optical remote sensing satellites, 48 low-orbit SAR satellites, and 56 electronic reconnaissance satellites, and the rolling time domain window length is set to 5 minutes; as Figures 1 to 3 shown, a contract net primary satellite selection method based on a mission area duty strategy includes the following steps:

[0173] S1. Obtain the space state data, platform state data, and mission state data of 56 satellites from the distributed satellite constellation within the first rolling time domain window; where:

[0174] The space state data is the ECEF space position and velocity of satellite i , where are the xyz-axis positions of satellite i in the ECEF system respectively, are the xyz-axis velocities of satellite i in the ECEF system respectively;

[0175] The platform state data is the satellite health of satellite i 、Remaining battery power 、Remaining memory 、Remaining bandwidth , denoted as ;

[0176] The task status data is the task completion rate of satellite i ;

[0177] S2. Respectively construct a set of space status data , a set of platform status data , and a set of task status data based on the space status data, platform status data, and task status data obtained in step S1; while constructing the set of space status data , set the initial number of settlements 6 , and set the minimum number of hops between satellites within a settlement 4 ; The set of platform status data includes a set of platform numbers ;

[0178] S3. Randomly select 6 coordinate points from the set of space status data as the initial settlement set , input the initial settlement set into the spatial density clustering algorithm, and set the number of iterations iter = 100 and the iteration accuracy error = 0.001 for iterative calculation to obtain the settlement set ;

[0179] S4. Define the link establishment constraint conditions for each satellite according to the geometric visibility constraint, inter-satellite distance constraint, and relative angular velocity constraint that affect the visibility between satellites. The link establishment constraint conditions are: geometric visibility constraint = inter-satellite distance constraint = relative angular velocity constraint = 1, that is , perform link establishment judgment on each satellite in the settlement, and let i = 1;

[0180] S5. Construct an undirected connected graph of satellites in the settlement of order , where is the vertex set of the connected graph, is the edge set of the connected graph, represents the edge from vertex v j to vertex v k ;

[0181] S6. According to the link establishment constraint conditions in step S4 and the connection matrix defined according to the undirected connected graph G in step S5, by performing operations on the settlement s iPerform connectivity judgment on all satellites in it and construct the connectivity matrix W i That is, the settlement s i 's routing table;

[0182] S7. Through the Dijkstra algorithm, repeatedly use the iterative formula:

[0183] ;

[0184] After iterative calculation, obtain the minimum hop count matrix mt i The minimum hop count matrix mt i 's expression is:

[0185] ;

[0186] In the formula, is the minimum hop count from vertex to vertex ;

[0187] According to the link establishment judgment results of each satellite in the settlement S, let i = 1, and iteratively calculate the settlement minimum hop count matrix set ; If the largest element in MT is less than the set minimum hop count of 4, that is , then output the settlement set S, otherwise N = N + 1, and return to step S3;

[0188] S8. Obtain the space state data set , the platform state data set , the task state data set and the clustering set S from steps S2 and S3 respectively, and input the ECEF coordinates of the task area;

[0189] S9. The settlement s i contains the platform number set , csp i,j is the number of the jth satellite in the settlement s i , record the center space coordinates of the settlement s i as , the distance dp i,j between the satellite csp and the settlement center j , there is a settlement s i center distance matrix ;

[0190] S10. Divide the motion period between satellites into a series of time slices with equal intervals based on the virtual topology strategy to construct a satellite network snapshot, and then calculate the satellite routing performance metrics, calculate the satellite platform performance metrics based on the satellite platform data set, and calculate the satellite mission performance metrics based on the satellite mission data set; the satellite routing performance metrics include the average satellite routing delay, satellite routing delay difference, average node degree, and node degree jitter; the satellite platform performance metrics include satellite health, remaining power, remaining memory, and remaining bandwidth; the satellite mission performance metric is the task backlog rate; where:

[0191] The matrix of the average routing delay is:

[0192] ;

[0193] In the formula, TS is the minimum delay matrix between satellites, is an all-1 matrix of m*1;

[0194] The matrix of the routing delay difference is:

[0195] ;

[0196] The matrix of the average node degree is:

[0197] ;

[0198] The matrix of the node degree jitter is:

[0199] ;

[0200] In the formula, represents taking the maximum value of each row of MW, represents taking the minimum value of each row of MW;

[0201] The matrix of the satellite health is:

[0202] ;

[0203] The matrix of the remaining power is:

[0204] ;

[0205] The matrix of the remaining memory is:

[0206] ;

[0207] The matrix of the remaining bandwidth is:

[0208] ;

[0209] The matrix of the task backlog rate is:

[0210] ;

[0211] Among them, is the total number of all tasks of satellite m, is the number of unfinished tasks of satellite m;

[0212] S11. According to the satellite routing performance index, satellite platform performance index and satellite task performance index in step S10, construct an index matrix , for settlement s i , substitute into the entropy weight method, and calculate the main star efficiency matrix i of each satellite in settlement s . Sort in descending order, and record the efficiency value ranked first as , and record the corresponding satellite number as . The efficiency value ranked second is , and record the corresponding satellite number as . And select the main star ms and the backup main star bs i from i according to the mission area on-duty strategy; where:

[0213] The mission area on-duty strategy is constructed based on the known efficiency matrix and the satellite number matrix ; The mission area on-duty strategy is:

[0214] a. If is located within the mission area Tg, is located outside the mission area Tg, then the main star ms i = , and the backup main star bs i = ;

[0215] b. If is located outside the mission area Tg, is located within the mission area Tg, and , then the main star ms i = , and the backup main star bs i = ;

[0216] c. If is located outside the mission area Tg, is located within the mission area Tg, and , then the main star ms i = , and the backup main star bsi = ;

[0217] d. If and are both outside the task area Tg, then the main satellite ms i = , and the backup main satellite bs i = ;

[0218] S12. Let i = 1, and iteratively calculate the main satellite set and the backup main satellite set based on step S11.

[0219] As Embodiment 4 of the present invention, the distributed satellite constellation is composed of satellites with various different payloads and different orbits; among them: the payloads include optical remote sensing payloads, SAR payloads, and electronic reconnaissance payloads; the orbit distribution covers low Earth orbit, medium Earth orbit, and geosynchronous orbit.

[0220] As Embodiment 5 of the present invention, the task scheduling of the distributed satellite constellation adopts an event-triggered hybrid scheduling framework to perform task scheduling on satellite resources within different rolling time domains in stages.

[0221] As Embodiment 6 of the present invention, the geometric visibility constraint requires that there is no direct Earth occlusion between satellites, and when any satellite is not in the atmosphere, it indicates that the satellites are geometrically visible. If the satellites are geometrically visible, then count = 1;

[0222] The inter-satellite distance constraint requires that the spatial distance between satellites is less than or equal to the maximum transmission distance, then the satellites satisfy the inter-satellite distance constraint, and count ;

[0223] The relative angular velocity constraint requires that the relative operation between satellites does not exceed the antenna acquisition and tracking capabilities of any satellite, then the satellites satisfy the relative angular velocity constraint, and count ran = 1.

[0224] As Embodiment 7 of the present invention, the process of constructing the routing table of the settlement s i :

[0225] The settlement s i contains satellites. Based on graph theory, construct an undirected connected graph of order , where is the vertex set of the connected graph, is the edge set of the connected graph, and represents the edge from vertex v to vertex v j to vertex v k ;

[0226] Based on the link establishment constraint conditions defined in step S4, the connectivity matrix of the undirected connected graph G is defined as:

[0227] ;

[0228] By performing connectivity judgment on all satellites in settlement s i to construct the connectivity matrix W i , which is the routing table of settlement s i .

[0229] As Embodiment 8 of the present invention, the satellite routing performance index is given by the minimum delay matrix between satellites; wherein: the calculation process of the minimum delay matrix includes the following steps:

[0230] S61. Based on graph theory, construct an undirected connected graph of all m satellites in the constellation , and calculate the distance l between satellite j and satellite k in the constellation j,k , then there is an inter-satellite distance matrix ;

[0231] S62. Based on the link establishment constraint conditions defined in step S4, construct the connectivity matrix and the delay matrix of the undirected connected graph G, and define as:

[0232] ,

[0233] where c is the speed of light;

[0234] S63. Through the Dijkstra algorithm, calculate the minimum delay matrix between satellites, denoted as:

[0235] ;

[0236] where is the minimum delay from vertex to vertex .

[0237] During the test, the low-earth orbit optical remote sensing satellite, the low-earth orbit SAR satellite, and the electronic reconnaissance satellite are all sun-synchronous circular orbit walker constellations with orbital altitudes of 520 km, 600 km, and 700 km respectively. The initial health of the satellites is a random value that conforms to the normal distribution between [0.7, 0.98]. The percentage numbers of the remaining power, remaining memory, and remaining bandwidth matrix are random values that conform to the normal distribution between [0.82, 0.96]. In Example 1, Example 2, and Example 3, the primary satellite selection method based on the minimum hop count priority is adopted. The average failure probabilities of the primary satellites of the distributed space-based system in the rolling time domain are 32.9%, 60.9%, and 60.6% respectively. The average failure probabilities calculated by the method of this patent are 28.8%, 56.1%, and 57.2% respectively. In the case of a single primary satellite failure, the cooperation efficiency of the distributed space-based system under the traditional primary satellite selection method is reduced by 12.1%, 15.5%, and 17.4% respectively, and the cooperation efficiency of the distributed space-based system under the primary satellite selection method based on the mission area guarding strategy is reduced by 4.2%, 8.2%, and 9.2% respectively. The primary satellite selection method of the distributed space-based system in the present invention effectively improves the reliability and robustness of the distributed space-based system.

Claims

1. A method for selecting a primary satellite for a contract network based on a mission area duty strategy, characterized in that: The steps include: S1. Acquire the space status data, platform status data and mission status data of m satellites from a distributed satellite constellation within a rolling time domain window; wherein: The space state data is the ECEF space position velocity of satellite i ,in, are the xyz three-axis positions of satellite i in the ECEF system, are the xyz three-axis velocities of satellite i in the ECEF system; The platform status data is the satellite health of satellite i , remaining power , Remaining memory , remaining bandwidth , recorded as ; The mission status data is the mission completion rate of satellite i ; S2: construct a space state data set based on the space state data, platform state data and task state data obtained in step S1. , platform status data collection , Task status data collection ; In constructing the spatial state data set At the same time, the initial number of settlements N is set, and the minimum number of hops M between satellites in the settlement is set; the platform status data set Includes platform number collection ; S3, according to the number of settlements N from the spatial state data set Randomly select N coordinate points as the initial settlement set , the initial settlement set Input the spatial density clustering algorithm, and set the number of iterations iter and the iteration accuracy error to perform iterative calculations to obtain the settlement set. ; S4. Define the constrained conditions for establishing a link for each satellite according to the geometric visible constraint, inter-satellite distance constraint and relative angular velocity constraint that affect the visibility between satellites. The constrained conditions for establishing a link are: geometric visible constraint = inter-satellite distance constraint = relative angular velocity constraint = 1, that is, , make a link judgment for each satellite in the settlement, let i=1; S5. Constructing the settlement based on graph theory Satellite Undirected connected graph ,in is the set of vertices of the connected graph, is the edge set of the connected graph, Represents vertex v j To the vertex v k The edge of S6: A connectivity matrix defined according to the link building constraint conditions in step S4 and the undirected connectivity graph G in step S5. , through the settlements i All satellites in the network are connected to determine the connectivity and construct the connectivity matrix W i , that is, settlements i Routing table; S7. Use the Dijkstra algorithm and repeatedly use the iterative formula: ； After iterative calculation, the minimum hop count matrix mt is obtained i , the minimum hop count matrix mt i The expression is: ； In the formula, From the vertex To the top The minimum number of hops; According to the link building judgment results of each satellite in the settlement S, let i=1, and iterate and calculate the minimum hop count matrix set of the settlement ; If the largest element in MT is less than the set minimum hop count M, that is , then output the settlement set S, otherwise N=N+1, and return to step S3; S8, respectively obtain the space state data set from steps S2 and S3 , platform status data collection , task status data set and the settlement set S, and enter the ECEF coordinates of the mission area ; S9, Settlements i Contains platform number collection ,csp i,j For settlements i The number of the jth satellite in the i The spatial coordinates of the center point are marked as , satellite csp i,j and settlement center Distance dp j , there are settlements i Center distance matrix ; S10, based on the virtual topology strategy, the movement cycle between satellites is divided into a series of equally spaced time slices to construct a satellite network snapshot, and then the satellite routing performance index is calculated, the satellite platform performance index is calculated based on the satellite platform data set, and the satellite mission performance index is calculated based on the satellite mission data set; the satellite routing performance index includes the satellite average routing delay, satellite routing delay difference, average node degree and node degree jitter; the satellite platform performance index includes satellite health, remaining power, remaining memory, and remaining bandwidth; the satellite mission performance index is the mission accumulation rate; wherein: The matrix of the average routing delay is: ； Where TS is the minimum delay matrix between satellites, is an m*1 matrix of all 1s; The matrix of the routing delay difference is: ； The matrix of the average node degree is: ； The matrix of the node degree jitter is: ； In the formula, It means taking the maximum value for each row of MW. It means taking the minimum value for each row of MW; The matrix of satellite health is: ； The matrix of the remaining power is: ； The matrix of the remaining memory is: ； The matrix of the remaining bandwidth is: ； The matrix of the task accumulation rate is: ； in, is the total number of missions of satellite m, is the number of unfinished missions of satellite m; S11, construct an indicator matrix according to the satellite routing performance indicators, satellite platform performance indicators and satellite mission performance indicators in step S10 , for settlements i ,Will Substituting into the entropy weight method, the settlement s is calculated i The main satellite efficiency matrix of each satellite in ,right Sort in descending order, and the performance value of the first performance ranking is recorded as , the corresponding satellite number is recorded as , the second most efficient value is , the corresponding satellite number is recorded as , and according to the mission area duty strategy Select the main star ms i and backup main star bs i ;in: The mission area guarding strategy is based on the known effectiveness matrix and satellite number matrix The mission area guard strategy is: a. If Located in the mission area Tg, If it is outside the mission area Tg, the main star ms i = , backup main star bs i = ; b. If Located outside the mission area Tg, is within the task area Tg, and , then the main star ms i = , backup main star bs i = ; c. If Located outside the mission area Tg, is within the task area Tg, and , then the main star ms i = , backup main star bs i = ; d. If and Both are located outside the mission area Tg, then the main star ms i = , backup main star bs i = ; S12, let i = 1, and iterate and calculate the primary star set based on step S11 and backup primary star set .

2. The method for selecting a primary satellite for a contract network based on a mission area duty strategy according to claim 1, characterized in that: The distributed satellite constellation is composed of satellites with different functional payloads and different orbits; wherein: the functional payloads include optical remote sensing payloads, SAR payloads, and electronic reconnaissance payloads; and the orbit distribution covers low orbit, medium orbit, and synchronous orbit.

3. The method for selecting a primary satellite for a contract network based on a mission area duty strategy according to claim 2 is characterized in that: The task scheduling of the distributed satellite constellation adopts an event-triggered hybrid scheduling framework to carry out task scheduling for satellite resources in different rolling time domains in stages.

4. The method for selecting a primary satellite for a contract network based on a mission area duty strategy according to claim 1, characterized in that: The geometric visibility constraint requires that there is no earth occlusion between satellites and that any satellite is not in the atmosphere. If the satellites are geometrically visible, then =1; The inter-satellite distance constraint requires that the spatial distance between satellites is less than or equal to the maximum transmission distance. If the inter-satellite distance constraint is satisfied between satellites, the calculation ; The relative angular velocity constraint requires that the relative movement between satellites does not exceed the antenna capture and tracking capability of any satellite. If the relative angular velocity constraint is satisfied between the satellites, ran=1.

5. The method for selecting a primary satellite for a contract network based on a mission area duty strategy according to claim 1, characterized in that: The settlements i The construction process of the routing table: Settlements i Included Satellites, built based on graph theory Undirected connected graph ,in is the set of vertices of the connected graph, is the edge set of the connected graph, Represents vertex v j To the vertex v k The edge of Based on the link building constraints defined in step S4, the connectivity matrix of the undirected connected graph G Defined as: ； Through the settlements i All satellites in the network are connected to determine the connectivity and build a connectivity matrix W i , that is, settlements i The routing table.

6. The method for selecting a primary satellite for a contract network based on a mission area duty strategy according to claim 1, characterized in that: The satellite routing performance index is given by the minimum delay matrix between satellites; wherein: the calculation process of the minimum delay matrix includes the following steps: S61. Construct an undirected connected graph of all m satellites in the constellation based on graph theory , and calculate the distance l between satellite j and satellite k in the constellation j,k , then the inter-satellite distance matrix ; S62: Based on the link building constraints defined in step S4, a connectivity matrix of the undirected connected graph G is constructed. and delay matrix , and Defined as: ， Where c is the speed of light; S63. Calculate the minimum delay matrix between satellites using the Dijkstra algorithm, expressed as: ； In the formula, From the vertex To the top Minimum delay.

Images