Multi-star cooperative planning method based on multi-decision main star
By employing a multi-decision primary star collaborative planning method, and utilizing multi-star computing resources for joint planning, the problems of high computational pressure and poor elasticity of a single node are solved, thus achieving efficient multi-star autonomous collaborative task management and resource scheduling.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- BEIJING INST OF CONTROL ENG
- Filing Date
- 2022-11-04
- Publication Date
- 2026-07-07
AI Technical Summary
Existing multi-satellite collaborative mission planning methods mainly rely on a single node, resulting in high computational pressure, insufficient utilization of multi-satellite computing resources, impacting the timeliness of mission allocation and exhibiting poor flexibility. They are also ill-suited to handle the random bursts and dynamic changes of multiple missions at multiple points in large constellations.
A collaborative planning method using multiple decision-making primary stars is adopted. Through parallel task planning, consistent pricing, task information synchronization, and distributed conflict resolution, the computing resources of multiple stars are used for joint planning, which distributes the planning pressure and improves the timeliness of task response and the flexibility of autonomous collaboration.
It has enabled multi-satellite autonomous collaborative task management, improved task response timeliness, reduced ground planning pressure, and enhanced the intelligence level of constellation autonomous collaboration and resilience to node failures.
Smart Images
Figure CN115756777B_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of multi-satellite intelligent autonomous collaborative control technology, and relates to a multi-satellite collaborative planning method based on multiple decision-making master stars. Background Technology
[0002] Large-scale constellations are a key future development direction. The US's "Flock" and SensorWeb constellations, among others, have been gradually deployed in orbit and put into application. Next-generation situational awareness constellations with high autonomous mission management capabilities, such as the Defense Space Architecture (NDSA) and the BlackJack constellation, are also under accelerated development and deployment, compensating for the limitations of single-satellite constellations in payload functionality, resolution, and spatiotemporal coverage. As space missions become increasingly complex, observation constellations are required to effectively cope with multi-mission, multi-point, random bursts, multi-mission resource scheduling conflicts, and dynamic changes in the temporal and spatial dimensions of observed targets. This places higher demands on the timeliness, continuity, and coordination of constellation observations. Traditional ground-based mission planning and control models are increasingly unable to meet these requirements, necessitating constellations to achieve autonomous mission management and planning scheduling through multi-satellite autonomous collaboration. However, current multi-satellite collaborative mission planning methods primarily focus on allocating multi-satellite tasks and resolving conflicts on a single node. This results in high computational burden on the single node, underutilization of multi-satellite computing resources, impacting the timeliness of task allocation, and exhibiting poor resilience, easily leading to global mission collapse upon the failure of a single node. Therefore, a multi-satellite distributed decision-making method is needed for large-scale observation constellations. This method should utilize multi-satellite computing resources for joint planning, distributing the pressure of parallel multi-task planning, improving mission response timeliness, and enhancing the constellation's resilience to the failure of a few nodes in its autonomous collaboration. Summary of the Invention
[0003] The technical problem solved by this invention is to overcome the shortcomings of existing technologies and propose a multi-satellite collaborative planning method based on multiple decision-making primary stars. From the perspective of engineering applications, this patent takes into account the needs of future constellation expansion, large number of tasks, and high timeliness requirements. For the multi-satellite distributed decision-making problem, it proposes a multi-satellite collaborative planning method. When multiple nodes of a remote sensing constellation receive observation tasks simultaneously, through parallel task planning of multiple decision-making primary stars, consistent pricing, task information synchronization, distributed conflict resolution decision-making, and global task joint decision-making, it can fully leverage the computing power of multiple stars to distribute the planning computing pressure, improve planning timeliness, reduce the amount of inter-satellite negotiation data and the number of interactions, and improve the efficiency of multi-satellite collaborative planning.
[0004] The solution of this invention is: a multi-star collaborative planning method based on multiple decision-making primary stars, comprising:
[0005] The constellation receives multiple burst missions, which trigger the formation of multiple decision-making primary stars according to the primary star generation rules. These multiple decision-making primary stars are responsible for different mission sets, and priority evaluation and parallel bidding are carried out. Candidate observation resources are determined based on the bidding information of the member stars.
[0006] The decision-making primary satellite sets a consistent price for candidate observation resources based on their observation quality, observation time, and execution cost.
[0007] The decision-making satellites will synchronize their respective task information, candidate observation resources, and corresponding prices across the entire network.
[0008] Based on the global task priority distribution and candidate resource pricing distribution, the multi-decision master star adopts a parameterized multi-task conflict resolution model to determine the conflict resolution parameters of each task according to tacit rules; it resolves conflicts for all network tasks and generates a resource pre-scheduling plan.
[0009] The multiple decision-making primary stars score their respective pre-scheduling schemes according to predetermined rules, and then synchronize the pre-scheduling schemes and their corresponding scores across the entire network, using the pre-scheduling scheme with the highest score as the final planning result.
[0010] The multiple primary decision-making satellites are responsible for different tasks, and priority evaluation and parallel bidding are conducted. Candidate observation resources are determined based on the bidding information of the member satellites, including:
[0011] Multiple decision stars S1, S2, S3, ... receive observation task sets A, B, C, ... respectively within the same planning period [t0, t0+T]; t0 represents the initial time, and T represents the interval time; the priority of the tasks is evaluated based on the user level LU of the uploaded tasks, the threat level LT of the observed targets, the time sensitivity characteristic LD, and the urgency level LE of the observations.
[0012] Pr=α u LU+α t LT+α d LD+α e LE
[0013] Where Pr represents the task priority, α u α t α d α e These represent priority weight coefficients for user level, threat level, time sensitivity, and urgency, respectively.
[0014] The decision-making primary star determines the visible satellites within the planning period [t0, t0+T] based on the constellation satellite orbits and the target's geographical location, and selects them as member stars.
[0015] The primary satellite, which makes the decision, issues task bidding information to the member satellites of the set of observation tasks under its jurisdiction, including task ID, priority Pr, target type, geographical location, imaging method, payload type, and resolution requirements.
[0016] Member satellites receive bidding tasks uniformly before and after, and determine whether the satellite's payload is feasible based on the type of payload, imaging method, and resolution requirements of the task.
[0017] If executable, the visible window for satellite imaging of the target is calculated using orbit prediction. Within this visible window, an imaging window that satisfies both the imaging time constraint and the attitude maneuver time constraint is searched. If an imaging window that satisfies the constraints exists, the mission executable flag Oi = 1, and the imaging attitude angle is calculated. Camera resolution OR, task completion time t os The energy consumption Op for executing the task is used as bidding information; if it is not executable, or although it is executable, there is no imaging window that satisfies the constraints, then the task executable flag Oi = 0; if a member star receives task bids from multiple decision master stars in the same period, no conflict resolution is performed, and the bidding information of each decision master star task is calculated independently.
[0018] Member stars send mission bidding information to the corresponding decision-making master stars.
[0019] The multi-decision master satellite performs consistent pricing based on the observation quality, observation time, and execution cost of the candidate observation resources. This includes: pricing each bidding scheme independently based on the bidding information of its respective task, according to the quality of observation, the time of completion, and the size of the task cost.
[0020] Pricing is based on the quality of the observation, specifically: the higher the imaging resolution, the higher the price.
[0021]
[0022] Where Meri1 represents the quality of observation, Or, ψ, θ These are the imaging resolution, imaging pitch angle, yaw angle, and roll angle from the bidding information of the member satellites.
[0023] Pricing is based on completion time, specifically: the earlier the imaging is completed, the higher the price.
[0024] Meri2=t os -t0
[0025] Meri2 represents the imaging time, t os t0 and t0 represent the time when the member star completes its observation mission and the current planned time, respectively.
[0026] Pricing is based on the cost of the task, specifically by the rule that the lower the energy consumption to complete the task, the higher the price.
[0027] Meri3=-Op
[0028] Meri3 represents the mission cost, and Op represents the energy required for the member star to perform attitude maneuvers during the observation mission.
[0029] Each bid proposal will be priced independently, specifically by the following method:
[0030] Meri=α1·Meri1+α2·Meri2+α3·Meri3
[0031] Meri represents the bid price of the proposed scheme, Meri1, Meri2, and Meri3 represent the observation quality, imaging time, and mission cost of the scheme, respectively, and α1, α2, and α3 are the weighting coefficients of the three factors.
[0032] The process of synchronizing mission information, candidate observation resources, and corresponding bids by the primary decision-making star includes: after completing an independent bid, the primary decision-making star synchronizes the mission ID, priority Pr mission information, and all corresponding bidding information with other primary decision-making stars in the constellation.
[0033] The multi-decision primary star employs a parameterized multi-task conflict resolution model and a candidate resource pricing distribution, determining its respective conflict resolution parameters according to tacit rules, including:
[0034] Based on the global task priority distribution, all task priorities are normalized and arranged from high to low. If any two candidate bids for tasks m1 and m2 have potential conflicts, the priority difference K = Pr between the two tasks is calculated. m1 -Pr m2 Where m1 has a higher priority than m2, forming a set KS of priority differences between tasks with potential conflicts;
[0035] Calculate the price difference L = Meri between candidate bids bid1 and bid2 that have resource scheduling conflicts. bid1 -Meri bid2 The bid price of bid1 is higher than that of bid2, forming a set LS of the differences in bid prices for tasks with potential resource scheduling conflicts;
[0036] Select v sets of parameters K and L from the set KS of priority differences and the set LS of bid price differences to generate a set of parameter K and L combinations.<K1,L1> , <K2,L2>,...,<K v ,L v >, as a parameter for multi-task conflict resolution, where v represents the number of combinations of parameters K and L;
[0037] After determining the parameters for resolving multi-task conflicts, the decision-making primary stars N within the constellation are used as the basis. s According to the principle of equal division, each decision-making star determines its own conflict resolution parameters. The conflict resolution parameters for decision-making star 1 are as follows:<K1,L1> , <K Ns+1 ,L Ns+1 >, <K 2Ns+1 ,L 2Ns+1 >,......;The conflict resolution parameters for decision-making primary star 2 are: <K2,L2>,<K Ns+2 ,L Ns+2 >, <K 2Ns+2 ,L 2Ns+2 >, ...; and so on.
[0038] The multi-decision primary star resolves conflicts among all network tasks and generates a resource pre-scheduling scheme, including:
[0039] S51. For each decision master star, for each set of K and L parameters, execute the following S52 to S54;
[0040] S52. Set all task scheduling flags to 0, and set all preset flags for all schemes to 0;
[0041] S53. For each task with a scheduling flag of 0, select bidding schemes in descending order of priority, and assign the highest bid with a pre-set flag of 0 as the observation resource for that task. If bid2 is selected for task m2 and the corresponding resource is not used by other tasks, then the scheduling flag of task m2 is set to 1, and the pre-set flag of the corresponding bid2 is set to 1. If bid2 is selected for task m2 and the corresponding resource has already been occupied by bid1 of task m1, calculate the priority difference between the two tasks, dPr = Pr. m2 -Pr m1 Calculate the price difference dMeri = Meri between the two conflicting bids. bid2 -Meri bid1 Conflict resource determination is performed based on conflict resolution parameters K and L; if there are no bids or no bids with a preset flag of 0 when selecting bids for task m2, the scheduling flag of task m2 is set to -1.
[0042] S54. Repeat S53 until all task scheduling flags are 1 or -1.
[0043] In step S53, the method for determining conflicting resources based on conflict resolution parameters K and L is as follows:
[0044] If dPr > K, then the resource is used by task m2, the scheduling flag of m2 is set to 1, and the corresponding bidding scheme preset flag is set to 1; the scheduling flag of m1 is set to 0, and the corresponding bidding scheme preset flag is set to -1.
[0045] If |dPr|≤K and dMeri>L, then the resource is used by task m2, the scheduling flag of m2 is set to 1, and the corresponding bidding scheme preset flag is set to 1; the scheduling flag of m1 is set to 0, and the corresponding bidding scheme preset flag is set to -1.
[0046] If |dPr|≤K, but dMeri≤L, then the resource is still used by task m1, the corresponding bidding scheme preset flag of m2 is set to -1, and other bidding schemes are selected.
[0047] If dPr < -K, the resource will still be used by task m1, the corresponding bidding scheme preset flag of m2 will be set to -1, and other bidding schemes will be selected.
[0048] The multi-decision primary satellites score their respective pre-scheduling schemes according to predetermined rules, then synchronize the pre-scheduling schemes and their corresponding scores across the entire network, using the pre-scheduling scheme with the highest score as the final planning result, including:
[0049] Based on all tasks m with scheduling flag set to 1 i The bid with a pre-set flag of 1 corresponds to the bid. j Calculate the global benefit of the scheduling scheme for each group of K and L.
[0050] M=∑Pr mi Meri bid_j
[0051] The multi-decision primary star synchronizes the scheme with the highest global benefit and its benefit among its respective pre-scheduling schemes; the pre-scheduling scheme with the highest global benefit is used as the final planning result.
[0052] The beneficial effects of this invention compared to the prior art are:
[0053] (1) Compared with the traditional multi-satellite collaborative mission planning method based on ground control, the present invention can realize multi-satellite autonomous collaborative mission management and resource scheduling, effectively cope with the mission scenarios such as multi-task multi-point random bursts, multi-task resource scheduling conflicts, and dynamic changes of observation targets in time / space of future large-scale constellations, meet the requirements of high timeliness, strong continuity, and complex collaborative mode proposed by the above-mentioned complex tasks, improve the intelligent autonomy of constellations, and significantly reduce the pressure of ground planning and control.
[0054] (2) Compared with the current proposed methods that focus on single-node multi-star task allocation and multi-task conflict resolution, this invention can effectively avoid the drawbacks of single-node planning under multi-task parallelism, such as high computational pressure, insufficient utilization of multi-star computing resources, impact on the timeliness of task allocation, and poor flexibility. It makes full use of multi-star computing resources for distributed decision-making, realizes joint planning, distributes the pressure of multi-task parallel planning, improves task response timeliness, and enhances the flexibility of constellation autonomous collaboration in the face of a few node failures. Attached Figure Description
[0055] Figure 1 This is a flowchart of a multi-star collaborative planning method based on multiple decision-making primary stars according to the present invention; Detailed Implementation
[0056] The present invention will be further described below with reference to the embodiments.
[0057] This invention proposes a multi-satellite collaborative planning method based on multiple decision-making primary stars. From the perspective of engineering applications, considering the future expansion of constellation scale, large number of tasks, and high timeliness requirements, this method addresses the distributed decision-making problem of multiple stars. When multiple nodes of a remote sensing constellation simultaneously receive observation tasks, this method fully leverages the computing power of multiple stars to achieve joint planning through distributed decision-making by using parallel priority evaluation, bidding, consistent pricing, task information synchronization, distributed multi-task conflict resolution, and global task joint decision-making. This effectively distributes the processing pressure of parallel multi-task planning and improves the efficiency of autonomous collaborative planning of multiple stars.
[0058] like Figure 1 As shown, the specific steps include the following:
[0059] Step 1: The constellation receives multiple burst missions and triggers the formation of multiple decision-making primary stars according to the primary star generation rules. The multiple decision-making primary stars are responsible for different mission sets. Priority evaluation and parallel bidding are carried out according to the following steps, and candidate observation resources are determined based on the bidding information of the member stars.
[0060] S1, multiple decision-making primary satellites S1, S2, S3, ..., within the same planning period [t0, t0+T], receive observation task sets A, B, C, ... respectively. Based on the user level LU of the assigned tasks, the threat level LT of the observed targets, the time sensitivity LD, and the urgency LE of the observations, the priority of the tasks is evaluated.
[0061] Pr=α u LU+α t LT+α d LD+α e LE (1)
[0062] Where Pr represents the task priority, αu α t α d α e These represent priority weight coefficients for user level, threat level, time sensitivity, and urgency, respectively.
[0063] S2. The decision-making master star obtains the visible satellites within the planning period window [t0, t0+T] based on the constellation satellite orbits and the target's geographical location, and uses them as member stars;
[0064] S3. The decision master satellite issues mission bidding information to the member satellites of the set of observation missions under its jurisdiction, including mission ID, priority Pr, target type, geographical location, imaging method, payload type, and resolution requirements.
[0065] S4. Member satellites receive bidding tasks uniformly before and after. Based on the payload type, imaging method, and resolution requirements of the task, they determine whether the satellite's payload is executable. If executable, they calculate the visible window through orbit prediction to image the target, and search within the visible window for imaging windows that meet the imaging time constraints and attitude maneuver time constraints. If an imaging window that meets the constraints exists, the task executable flag Oi = 1, and the imaging attitude angle is calculated. Camera resolution OR, task completion time t os The energy consumption Op for executing the task is used as bidding information; otherwise, the task executable flag Oi = 0. If a member star receives task bids from multiple decision master stars in the same period, no conflict resolution is performed, and the bidding information is calculated independently for each.
[0066] S5, member stars send task bidding information to the corresponding decision master star.
[0067] Step 2: The decision-making primary star, based on the bidding information for each mission, independently prices all bidding proposals according to the quality of observations, the time of completion, and the cost of the mission.
[0068] The method for pricing based on observation quality is to price according to the rule that higher imaging resolution results in a higher price.
[0069]
[0070] Where Meri1 represents the quality of observation, Or, ψ, θ These are the imaging resolution, imaging pitch angle, yaw angle, and roll angle from the bidding information of the member satellites.
[0071] The pricing method based on completion time is as follows: the earlier the imaging is completed, the higher the price.
[0072] Meri2=t os -t0 (3)
[0073] Meri2 represents the imaging time, t os t0 and t0 represent the time when the member star completes its observation mission and the current planned time, respectively.
[0074] The method for pricing based on task cost is to price the task according to the rule that the lower the energy consumption to complete the task, the higher the price.
[0075] Meri3=-Op (4)
[0076] Meri3 represents the mission cost, and Op represents the energy required for the member star to perform attitude maneuvers during the observation mission.
[0077] Each bidding proposal will be priced independently based on its observation quality, completion time, and mission cost.
[0078] Meri=α1·Meri1+α2·Meri2+α3·Meri3 (5)
[0079] Meri represents the bid price of the proposed scheme, Meri1, Meri2, and Meri3 represent the observation quality, imaging time, and mission cost of the scheme, respectively, and α1, α2, and α3 are the weighting coefficients of the three factors.
[0080] Step 3: After the primary satellite completes its independent bidding, it will synchronize its respective task information, candidate observation resources, and corresponding bids across the entire network, including task ID, priority Pr, and all relevant bidding information.
[0081] Step 4: Based on the global task priority distribution and candidate resource pricing distribution, the multi-decision master star adopts a parameterized multi-task conflict resolution model, determines the conflict resolution parameters of each task according to the following rules, resolves conflicts for all network tasks, and generates a resource pre-scheduling plan:
[0082] S1. Normalize the priorities of all tasks according to the global task priority distribution and arrange them from high to low. If any two candidate bids for tasks m1 and m2 have potential conflicts, calculate the priority difference K = Pr between the two tasks. m1 -Pr m2 Where m1 has a higher priority than m2, a set KS is formed representing the priority differences between potentially conflicting tasks.
[0083] S2. Calculate the price difference dMeri = Meri between candidate bids bid1 and bid2 that have resource scheduling conflicts. bid1 -Meri bid2 The bid price of bid1 is higher than that of bid2, forming a set LS of the differences in bid prices for tasks with potential resource scheduling conflicts;
[0084] S3. Select an appropriate number of parameters K and L from the set KS of priority differences and the set LS of bid price differences to generate a combination set of parameters K and L.<K1,L1> ,<K2,L2> ,..., <K v ,L v >, as a parameter for multi-task conflict resolution, where v represents the number of combinations of parameters K and L;
[0085] S4. After determining the parameters for resolving multi-task conflicts, based on the number N of the primary decision stars within the constellation. s According to the principle of equal division, each decision-making star determines its own conflict resolution parameters. The conflict resolution parameters for decision-making star 1 are as follows:<K1,L1> , <K Ns+1 ,L Ns+1 >, <K 2Ns+1 ,L 2Ns+1 >,......;The conflict resolution parameters for decision primary star 2 are:<K2,L2> , <K Ns+2 ,L Ns+2 >, <K 2Ns+2 ,L 2Ns+2 >,......; and so on.
[0086] S5. For each decision master star, for each set of K and L parameters, execute the following S6 to S8;
[0087] S6. Set all task scheduling flags to 0, and set all preset flags for all schemes to 0;
[0088] S7. For each task with scheduling flag 0, select bidding schemes in order of priority from high to low, and arrange the highest bid with the preset flag 0 as the observation resource for the task as much as possible. After each task completes the bidding, the corresponding scheduling flag is set to 1, and the corresponding preset flag is set to 1 after each bid is selected.
[0089] If bid2 is selected for task m2, and the corresponding resource is already occupied by bid1 for task m1, calculate the priority difference between the two tasks: dPr = Pr. m2 -Pr m1 Calculate the price difference dMeri = Meri between the two conflicting bids. bid2 -Meri bid1The conflict resource determination is performed based on the conflict resolution parameters K and L: If dPr > K, or dPr ≤ K and dMeri > L, then the resource is used by task m2, m2's scheduling flag is set to 1, and the corresponding bidding scheme preset flag is set to 1; m1's scheduling flag is set to 0, and the corresponding bidding scheme preset flag is set to -1; if dPr ≤ K and dMeri ≤ L, or dPr < -K, then the resource is still used by task m1, m2's corresponding bidding scheme preset flag is set to -1, and other bidding schemes are selected; if a task finds that all the preset flags of the bids are already -1 when selecting a bid, then the scheduling flag is set to -1;
[0090] S9. Repeat S7 above until all task scheduling flags are 1 or -1.
[0091] Step 5: The multi-decision primary satellites score their respective pre-scheduling schemes according to predetermined rules, then synchronize the pre-scheduling schemes and corresponding scores across the entire network, and use the pre-scheduling scheme with the highest score as the final planning result.
[0092] S1, based on all tasks m whose scheduling flag is 1. i The bid with a pre-set flag of 1 corresponds to the bid. j Calculate the global benefit of the scheduling scheme for each group of K and L.
[0093] M=∑Pr mi Meri bid_j (6)
[0094] S2. The multi-decision primary star will synchronize the scheme with the highest global return and its return among its respective pre-scheduling schemes.
[0095] S3, the multi-decision primary star uses the pre-scheduling scheme with the highest global benefit as the final planning result.
[0096] Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make possible changes and modifications to the technical solutions of the present invention by utilizing the methods and techniques disclosed above without departing from the spirit and scope of the present invention. Therefore, any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solutions of the present invention shall fall within the protection scope of the technical solutions of the present invention.
Claims
1. A multi-star collaborative planning method based on multiple decision-making primary stars, characterized in that... include: The constellation receives multiple burst missions, which trigger the formation of multiple decision-making primary stars according to the primary star generation rules. These multiple decision-making primary stars are responsible for different mission sets, and priority evaluation and parallel bidding are carried out. Candidate observation resources are determined based on the bidding information of the member stars. The decision-making primary satellite sets a consistent price for candidate observation resources based on their observation quality, observation time, and execution cost. The decision-making satellites will synchronize their respective task information, candidate observation resources, and corresponding prices across the entire network. Based on the global task priority distribution and candidate resource pricing distribution, the multi-decision master star adopts a parameterized multi-task conflict resolution model and determines the conflict resolution parameters of each task according to the tacit rules. Conflict resolution is performed on all tasks across the network to generate a resource pre-scheduling plan; The multi-decision primary star employs a parameterized multi-task conflict resolution model and a candidate resource pricing distribution, determining its respective conflict resolution parameters according to tacit rules, including: Based on the global task priority distribution, all task priorities are normalized and arranged from high to low. If any two tasks... and If there is a potential conflict between candidate bids, the priority difference between the two tasks should be calculated separately. ,in Higher priority This forms a set of priorities differences between tasks that have potential conflicts. ; Calculate candidate bids with resource scheduling conflicts and The difference in the listed price ,in The listed price is higher than This forms a set of bid price differences for tasks with potential resource scheduling conflicts. ; In the set of the aforementioned priority differences The set of differences from the bid price Selected from Group parameters and Generate parameters and Combination set As a parameter for multi-task conflict resolution, among which Representative parameters and The number of combinations; After determining the parameters for resolving multi-task conflicts, the decision-making primary stars within the constellation are used as a basis. According to the principle of equal division, each decision-making star determines its own conflict resolution parameters. The conflict resolution parameters for decision-making star 1 are as follows: The conflict resolution parameters for Decision Master Star 2 are: And so on; The multiple decision-making primary stars score their respective pre-scheduling schemes according to predetermined rules, and then synchronize the pre-scheduling schemes and their corresponding scores across the entire network, using the pre-scheduling scheme with the highest score as the final planning result.
2. The multi-star collaborative planning method based on multiple decision-making primary stars according to claim 1, characterized in that: The multiple primary decision-making satellites are responsible for different tasks, and priority evaluation and parallel bidding are conducted. Candidate observation resources are determined based on the bidding information of the member satellites, including: Multiple decision stars S1, S2, S3, ..., within the same planning cycle Within, observation task sets A, B, C, ... were received respectively; Indicates the initial time. Indicates the interval time; based on the user level of the above tasks. Threat level of the observed target Time-sensitive characteristics And the urgency of observation Prioritize the evaluation tasks in, Represents task priority. , , , These represent priority weight coefficients for user level, threat level, time sensitivity, and urgency, respectively. The decision-making primary star determines the planning cycle based on the constellation satellite orbits and the target's geographical location. Visible satellites within the constellation, serving as member satellites; The primary satellite, which makes the decision-making process, issues mission bidding information to its member satellites, including mission details. Priority Target type, geographical location, imaging method, payload type, and resolution requirements; Member satellites receive bidding tasks uniformly before and after, and determine whether the satellite's payload is feasible based on the type of payload, imaging method, and resolution requirements of the task. If executable, the visible window for satellite imaging of the target is calculated using orbit prediction. Within this visible window, an imaging window that satisfies both the imaging time constraint and the attitude maneuver time constraint is searched. If an imaging window that satisfies the constraints exists, the mission is considered executable. Calculate the imaging attitude angle Camera resolution Task completion time Energy consumption for performing this task This serves as bidding information; if it is not executable, or although executable, there is no imaging window that satisfies the constraints, then the task is executable. If a member star receives task bids from multiple decision-making master stars simultaneously within the same period, no conflict resolution is performed, and the bid information for each decision-making master star's task is calculated independently. Member stars send mission bidding information to the corresponding decision-making master stars.
3. The multi-star collaborative planning method based on multiple decision-making primary stars according to claim 1, characterized in that: The multi-decision master satellite performs consistent pricing based on the observation quality, observation time, and execution cost of the candidate observation resources. This includes: pricing each bidding scheme independently based on the bidding information of its respective task, according to the quality of observation, the time of completion, and the size of the task cost.
4. The multi-star collaborative planning method based on multiple decision-making primary stars according to claim 3, characterized in that: Pricing is based on the quality of the observation, specifically: the higher the imaging resolution, the higher the price. in This represents the quality of the observation. , , , These are the imaging resolution, imaging pitch angle, yaw angle, and roll angle from the bidding information of the member satellites.
5. The multi-star collaborative planning method based on multiple decision-making primary stars according to claim 3, characterized in that: Pricing is based on completion time, specifically: the earlier the imaging is completed, the higher the price. in This indicates the time of imaging. and These represent the completion time of the observation mission for each member satellite and the current planned time.
6. The multi-star collaborative planning method based on multiple decision-making primary stars according to claim 3, characterized in that: Pricing is based on the cost of the task, specifically by the rule that the lower the energy consumption to complete the task, the higher the price. in This represents the size of the task cost. The energy consumed to perform attitude maneuvers for member stars to carry out observation missions.
7. The multi-star collaborative planning method based on multiple decision-making primary stars according to claim 3, characterized in that: Each bid proposal will be priced independently, specifically by the following method: in The price quoted for the bid proposal. , , The factors considered are the observation quality, imaging time, and mission cost of the proposed scheme. , , These are the weighting coefficients of the three factors.
8. The multi-star collaborative planning method based on multiple decision-making primary stars according to claim 1, characterized in that: The synchronization of mission information, candidate observation resources, and corresponding prices by the primary decision-making star includes: after completing independent pricing, the primary decision-making star synchronizes the mission information with other primary decision-making stars in the constellation. Priority Task information, and all corresponding bidding information.
9. The multi-star collaborative planning method based on multiple decision-making primary stars according to claim 1, characterized in that: The multi-decision primary star resolves conflicts among all network tasks and generates a resource pre-scheduling scheme, including: S51. For each decision master star, for each set of K and L parameters, execute the following steps S52 to S54; S52. Set all task scheduling flags to 0, and set all preset flags for all schemes to 0; S53. For each task with a scheduling flag of 0, select bidding schemes in descending order of priority, and allocate the highest bid with a pre-set flag of 0 as the observation resource for the task. If it is a task... Selection If the corresponding resources are not used by other tasks, then the task... The scheduling flag is set to 1, and the corresponding bid pre-setting flag is set. Set to 1; if it is a task Selection At that time, the corresponding resources have been assigned to the task. Selection of targets Occupation, calculating the priority difference between two tasks. Calculate the price difference between the two conflicting bids. Conflict resources are determined based on conflict resolution parameters K and L; if it is a task If there are no bids during the bidding process, or if there are no bids with a pre-set flag of 0, then the task... Set the scheduling flag to -1; S54. Repeat S53 until all task scheduling flags are 1 or -1.
10. A multi-star collaborative planning method based on multiple decision-making primary stars according to claim 9, characterized in that: In step S53, the method for determining conflicting resources based on conflict resolution parameters K and L is as follows: if Then the resource is provided by the task. use, The scheduling flag is set to 1, and the corresponding bidding scheme preset flag is set to 1; Set the scheduling flag to 0 and the corresponding bidding scheme preset flag to -1; if ,and Then the resource is provided by the task. use, The scheduling flag is set to 1, and the corresponding bidding scheme preset flag is set to 1; Set the scheduling flag to 0 and the corresponding bidding scheme preset flag to -1; if ,but Then the resource is still provided by the task. use, Set the corresponding bidding scheme's pre-set flag to -1, and select other bidding schemes; if Then the resource is still provided by the task. use, Set the corresponding bidding scheme's pre-set flag to -1, and select other bidding schemes.
11. A multi-star collaborative planning method based on multiple decision-making primary stars according to claim 10, characterized in that: The multi-decision primary satellites score their respective pre-scheduling schemes according to predetermined rules, then synchronize the pre-scheduling schemes and their corresponding scores across the entire network, using the pre-scheduling scheme with the highest score as the final planning result, including: Based on all tasks with scheduling flag set to 1 Bids with a pre-set flag of 1 corresponding to the bid Calculate each group and The corresponding scheduling scheme has global benefits. The multi-decision primary star synchronizes the scheme with the highest global benefit and its benefit among its respective pre-scheduling schemes; the pre-scheduling scheme with the highest global benefit is used as the final planning result.