Constellation measurement and control task planning method and device based on dynamic graph neural network

Abstract

The invention relates to the technical field of aerospace mission planning, and discloses a constellation measurement and control mission planning method and device based on a dynamic graph neural network, and the method comprises the steps: obtaining a measurement and control mission set in a planning period from the measurement and control demands of a user, and determining a related satellite; obtaining a visible time window of each satellite to each antenna of the ground station, and performing time sequence arrangement on a starting time point and an ending time point of each visible time window to obtain a time step set; based on the node set and the edge set of each time step of the time step set, utilizing a dynamic graph neural network model to obtain a link combination set of each time step; and determining a plurality of target constellation measurement and control tasks from the measurement and control task set according to all the link combination sets, and determining the target constellation measurement and control tasks meeting constraint conditions as executable constellation measurement and control tasks to obtain a planning result. According to the method, the constellation measurement and control task planning efficiency can be improved, and the operation benefit and the coping capacity of a constellation system in a complex space environment are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of space mission planning, and particularly relates to a constellation TT&C mission planning method and device based on a dynamic graph neural network. BACKGROUND

[0002] A constellation system is a cooperative network composed of multiple artificial satellites according to specific orbital parameters, spatial configurations and functional designs, and achieves complex space mission objectives that a single satellite cannot achieve through satellite division and cooperation, and solves the deficiencies of traditional single-satellite systems in terms of coverage range, response speed, data redundancy and system robustness.

[0003] With the rapid development of satellite constellations, global connectivity is currently being reshaped, and countries are racing to seize space resources and consolidate their competitive advantage in space. Constellation TT&C mission planning refers to, in space engineering, planning an effective satellite-ground connection scheme for a constellation system composed of multiple satellites, taking into account constellation configuration, task priority, inter-satellite cooperation and resource constraints, etc., to maximize the mission effectiveness of the constellation system and optimize resource utilization, which requires the algorithm to quickly respond to dynamic changes in the space environment and task requirements. Currently, the research on constellation TT&C mission planning is relatively blank, and the common methods include classical mathematical model method, heuristic search algorithm, reinforcement learning method, etc., but these methods perform poorly when dealing with large-scale task sets, and cannot adapt to the requirements of dynamic changes in the space environment and task requirements. In the existing technology of constellation TT&C mission planning: The invention patent with publication number "CN116681235A" discloses a spaceborne distributed constellation cooperative autonomous mission planning system and method, which mainly includes a cooperative planning unit, a single-satellite planning unit and a task execution unit. The cooperative planning unit is responsible for receiving external tasks, analyzing task requirements and confirming task types, performing multi-satellite task allocation and conflict resolution according to benefit maximization and task priority, and generating a task scheduling scheme for each participating task satellite. The single-satellite planning unit receives the task scheduling scheme and performs task feasibility judgment, matches the task template on the satellite to generate task template instructions. The task execution unit receives the task template instructions, generates satellite executable instruction chains and outputs control instructions for the on-board single machine, and simultaneously collects task state information. However, this planning method is mainly based on the traditional task planning framework, and has low computational efficiency when dealing with large-scale task sets, cannot meet the real-time requirements, and has limited ability to quickly adjust in the face of unexpected events, and has deficiencies in dynamic environment adaptability and real-time performance.

[0004] The invention patent with the publication number "CN113327028A" discloses a constellation satellite task planning method, device and storage medium, mainly relates to obtaining a task planning search tree, and determining the task planning information of each satellite in the constellation by the shortest search path, determining the task planning rate based on the task planning information, and adding the task planning information to the task planning template when the task planning rate meets the judgment condition. This method establishes a task planning search tree, combines satellite task priority and measurement and control mode for task planning, and is suitable for the case where the data volume of the constellation is relatively small. However, this method is a planning method based on search tree, and also has the problem of insufficient adaptability to dynamic change environment, and has limitations in large-scale task planning and real-time performance, and lacks rapid response capability and high efficient computing performance in complex dynamic environment.

[0005] Therefore, there is an urgent need to provide a technical solution to solve the above problems. SUMMARY

[0006] To solve the above technical problems, the present application provides a constellation measurement and control task planning method and device based on dynamic graph neural network.

[0007] In a first aspect, the present application provides a constellation measurement and control task planning method based on dynamic graph neural network, and the technical solution of the method is as follows: Obtain a set of to-be-planned constellation measurement and control tasks in a target planning period from user measurement and control requirements, and determine the target satellites involved in the set of to-be-planned constellation measurement and control tasks; Obtain the visible time window of each target satellite to each target antenna of the ground station to form a set of visible time windows, and arrange the start time point and the end time point corresponding to each visible time window in the set of visible time windows in time sequence to obtain a set of target time steps; Based on the node set and edge set corresponding to each time step of the set of target time steps, and using the trained dynamic graph neural network model, obtain a set of target link combinations corresponding to each time step in the set of target time steps; According to all the sets of target link combinations, determine a plurality of target constellation measurement and control tasks from the set of to-be-planned constellation measurement and control tasks, and determine the target constellation measurement and control tasks that meet the constraint conditions as executable constellation measurement and control tasks to form the constellation measurement and control task planning result of the target planning period.

[0008] The beneficial effects of the constellation measurement and control task planning method based on dynamic graph neural network of the present application are as follows: The method of the present application optimizes constellation TT&C task planning through a dynamic graph neural network, can efficiently process a large-scale task set, can quickly respond to dynamic changes in the space environment and task requirements, solves the shortcomings of the existing planning mode in dynamic adaptability and real-time performance, improves the efficiency and effect of constellation TT&C task planning, and improves the operation benefit and response capability of the constellation system in a complex space environment.

[0009] On the basis of the above-mentioned scheme, the constellation TT&C task planning method based on the dynamic graph neural network of the present application can be further improved as follows.

[0010] In an optional manner, the step of acquiring the visible time window of each target antenna of the ground station by any target satellite comprises: Based on the overflight prediction information of the any target satellite, at least one visible time window of each target antenna of the ground station by the any target satellite is determined.

[0011] The beneficial effects of the above-optional manner are that the satellite overflight prediction information is further utilized to acquire the visible time window, the communication time range of the satellite and the ground station antenna can be accurately determined, accurate time basis is provided for subsequent task planning, and the timeliness and feasibility of the task planning are enhanced.

[0012] In an optional manner, the step of constructing the node set corresponding to any time step in the target time step set comprises: Based on all target antennas of the ground station and all target satellites visible at the any time step, the node set corresponding to the any time step is constructed.

[0013] The beneficial effects of the above-optional manner are that the node set is further constructed based on the ground station antenna and the visible satellite, the available resource nodes of each time step can be clearly presented, accurate input is provided for the dynamic graph neural network model, and accurate prediction of the target link combination is facilitated.

[0014] In an optional manner, the step of constructing the edge set corresponding to any time step in the target time step set comprises: Based on the node set corresponding to the any time step, and in combination with the target satellite and the target antenna having a link relationship at the any time step, the edge set corresponding to the any time step is constructed.

[0015] The beneficial effects of the above-optional manner are that the edge set is further constructed in combination with the node set and the satellite antenna link relationship, the connection between nodes is accurately described, the graph neural network model can better capture the interaction information between nodes, and the rationality of the task planning is improved.

[0016] In an optional manner, When the number of target satellites is greater than or equal to the number of target antennas among the target satellites and the target antennas in the link relationship at the any time step, the target link combination set corresponding to the any time step comprises: the antenna-satellite link combination with the highest link probability corresponding to each target antenna in the link relationship at the any time step. When the number of target satellites is less than the number of target antennas among the target satellites and the target antennas in the link relationship at the any time step, the target link combination set corresponding to the any time step comprises: the antenna-satellite link combination with the highest link probability corresponding to each target satellite in the link relationship at the any time step.

[0017] The beneficial effect of the above optional mode is that the target link combination is further determined according to the number relationship between the satellite and the antenna, the communication resources can be effectively allocated, the optimal link scheme can be selected under different resource proportion conditions, and the resource utilization rate and the task execution success rate are improved.

[0018] In an optional mode, the step of determining a plurality of target constellation TT&C tasks from the set of to-be-planned constellation TT&C tasks according to the set of all target link combinations comprises: For each to-be-planned constellation TT&C task in the set of to-be-planned constellation TT&C tasks, the to-be-planned constellation TT&C task that meets the preset condition is determined as a target constellation TT&C task; wherein the preset condition is that there is a target time window containing the to-be-measured time window of the to-be-planned constellation TT&C task, and the preset link combination corresponding to the to-be-planned constellation TT&C task is the same as the target link combination corresponding to the start and end time of the target time window.

[0019] The beneficial effect of the above optional mode is that the target constellation TT&C task is further screened through the preset condition, the required task can be quickly locked, the redundancy and error options in task planning are reduced, and the accuracy and efficiency of task planning are improved.

[0020] In an optional mode, the constraint condition is: The length of the to-be-measured time window of the target constellation TT&C task is not less than the target length; The to-be-measured time window of the target constellation TT&C task belongs to any one of the set of visible time windows; When the to-be-measured time window of the target constellation TT&C task has an overlapping time window, the target constellation TT&C task is the task with the highest preset priority in the overlapping task group to which the target constellation TT&C task belongs; wherein all target constellation TT&C tasks in the overlapping task group have the same time window, and the target satellite and / or target antenna in the preset link combination corresponding to each target constellation TT&C task in the overlapping task group are the same; The preset link combination of the target constellation TT&C task is any one of all fixed link combinations.

[0021] The beneficial effect of the optional mode is that the multi-dimensional constraint conditions are further defined to ensure the feasibility and rationality of the task, guarantee the key elements such as task duration and priority, avoid resource conflicts and task overlaps, and make the task planning more scientific and effective.

[0022] In a second aspect, the present application provides a constellation TT&C task planning device based on a dynamic graph neural network, and the technical scheme of the device is as follows: It comprises a determination module, a construction module, a prediction module and a planning module. The determination module is configured to obtain a set of to-be-planned constellation TT&C tasks in a target planning period from user TT&C requirements, and determine target satellites involved in the set of to-be-planned constellation TT&C tasks. The construction module is configured to obtain a visible time window of each target antenna of each target satellite to a ground station to form a set of visible time windows, and arrange a starting time point and an ending time point corresponding to each visible time window in the set of visible time windows in time sequence to obtain a set of target time steps. The prediction module is configured to obtain a set of target link combinations corresponding to each time step in the set of target time steps based on a set of nodes and a set of edges corresponding to each time step in the set of target time steps, and utilize a trained dynamic graph neural network model. The planning module is configured to determine a plurality of target constellation TT&C tasks from the set of to-be-planned constellation TT&C tasks according to all sets of target link combinations, and determine target constellation TT&C tasks meeting constraint conditions as executable constellation TT&C tasks to form a constellation TT&C task planning result of the target planning period.

[0023] The constellation TT&C task planning device based on a dynamic graph neural network has the following beneficial effects: The device optimizes constellation TT&C task planning through a dynamic graph neural network, can efficiently process a large-scale task set, can quickly respond to dynamic changes in the space environment and task requirements, solves the deficiencies of the existing planning mode in dynamic adaptability and real-time performance, improves the efficiency and effectiveness of constellation TT&C task planning, and improves the operation efficiency and response capability of the constellation system in a complex space environment.

[0024] In a third aspect, the technical scheme of an electronic device of the present application is as follows: It comprises a memory, a processor and a program stored on the memory and running on the processor, and the processor implements the steps of the constellation TT&C task planning method based on a dynamic graph neural network of the present application when executing the program.

[0025] In a fourth aspect, the present application provides a computer readable storage medium, and the technical solution is as follows: The computer readable storage medium stores instructions, and when the computer readable storage medium reads the instructions, the computer readable storage medium executes the steps of the constellation TT&C task planning method based on a dynamic graph neural network.

[0026] The above description is only a summary of the technical solutions of the present application. In order to more clearly understand the technical means of the present application, the content of the specification can be implemented, and in order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the following specific embodiments of the present application are described. BRIEF DESCRIPTION OF DRAWINGS

[0027] The accompanying drawings are only used to show the embodiments and are not considered as limitations of the present application. Moreover, the same reference signs are used to represent the same parts throughout the drawings. In the drawings: Figure 1 The flowchart of an embodiment of a constellation TT&C task planning method based on a dynamic graph neural network of the present application; Figure 2 The structural diagram of a constellation system; Figure 3 The random walk diagram of Node2Vec; Figure 4 The structural diagram of an embodiment of a constellation TT&C task planning device based on a dynamic graph neural network of the present application; Figure 5 The structural diagram of an embodiment of an electronic device of the present application. DETAILED DESCRIPTION

[0028] Exemplary embodiments of the present application will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present application are shown in the drawings, it should be understood that the present application can be implemented in various forms and should not be limited by the embodiments described herein.

[0029] Figure 1A flowchart of an embodiment of a constellation TT&C task planning method based on a dynamic graph neural network provided by the present application is shown, which can be executed by terminal equipment or a server, etc. electronic device. Among them, the terminal equipment can be any fixed or mobile terminal such as user equipment (User Equipment, UE), mobile equipment, user terminal, terminal, cellular phone, cordless phone, personal digital assistant (Personal Digital Assistant, PDA), handheld device, computing device, vehicle-mounted device, wearable device, etc. The server can be a single server or a server cluster composed of multiple servers. Any electronic device can realize the constellation TT&C task planning method based on a dynamic graph neural network by calling the computer readable instructions stored in the memory through the processor. As shown in Figure 1 , including the following steps: S1, obtaining a set of to-be-planned constellation TT&C tasks in a target planning period from user TT&C requirements, and determining target satellites involved in the set of to-be-planned constellation TT&C tasks.

[0030] Among them, the user TT&C requirement is the TT&C task request proposed by the scientific user, which contains the satellite TT&C task type to be executed, the preset link (antenna and satellite) combination, the task time window requirement and the task benefit level (divided into core task, important task and general task, corresponding to different priorities). The target planning period is the time range of task planning, expressed as . is the start time of the target planning period, is the end time of the target planning period.

[0031] Among them, the set of to-be-planned constellation TT&C tasks is the set of tasks that need to be planned in the target planning period . represents the number of to-be-planned constellation TT&C tasks in the target planning period; the th to-be-planned constellation TT&C task in the target planning period represents that the satellite needs to perform TT&C operation through the antenna in the time window , is the task benefit value (preset priority, the larger the value, the higher the priority).

[0032] Among them, the target satellite is all satellites involved in the set of to-be-planned constellation TT&C tasks, which constitutes a target satellite set A ground station is a ground measurement and control facility with multiple antennas, which is used to receive plans from the mission support center and perform satellite command injection and data transmission tasks. A target antenna set is a set of all antennas used by a ground station to perform measurement and control tasks. , one antenna can only serve one satellite at a time.

[0033] It should be noted that if Figure 2 As shown, the constellation system in this embodiment primarily consists of a ground-based measurement and control network and in-orbit satellites. The ground-based measurement and control network includes ground stations, a mission support center, and scientific users. Ground stations typically have multiple antennas in different frequency bands, but each antenna can only perform measurement and control tasks for one satellite at a time. The mission support center collects the measurement and control requirements of scientific users, generates work plans for the corresponding antennas at each ground station through resource constraint checks and conflict resolution, and transmits these plans to each ground station. The ground station receives and analyzes the work plans, and rotates the antennas according to the planned time and the corresponding satellite guidance data to perform satellite command injection and data transmission tasks. In-orbit satellites operate according to a predetermined orbital period and need to perform inter-satellite, satellite-to-ground measurements, and other in-orbit scientific experiments within the visible range of the ground station.

[0034] S2. Obtain the visible time window of each target satellite to each target antenna of the ground station to form a visible time window set, and arrange the start time point and the end time point corresponding to each visible time window in the visible time window set in time sequence to obtain a target time step set.

[0035] Among them, the visible time window refers to the target satellite Target antenna of ground station Visible communication period , determined according to the satellite passing forecast, Target satellite Target antenna No. The start time of the visible time window, Target satellite Target antenna No. The end time of the visible time window. It should be noted that the number of visible time windows of a target satellite to a target antenna is at least one ( The visible time window set W , Indicates the target satellite Target antenna The number of visible time windows, i ranges from [1,m], and j ranges from [1,n].

[0036] The target time step set refers to a set of discrete time points arranged in time sequence according to the start time points and end time points of all visible time windows .

[0037] S3, based on the node set and edge set corresponding to each time step of the target time step set, and using the trained dynamic graph neural network model, obtaining a target link combination set corresponding to each time step in the target time step set.

[0038] Each time step corresponds to a node set and an edge set. The dynamic graph neural network model refers to a graph neural network model trained to generate node embedding vectors using Node2Vec, aggregate historical states through a time series graph convolution layer, update node states through a gating mechanism, and output the link probability of the target satellite and the target antenna.

[0039] S4, from the target link combination set, determine a plurality of target constellation control tasks from the set of to-be-planned constellation control tasks, and determine the target constellation control task meeting the constraint condition as an executable constellation control task to form the constellation control task planning result of the target planning period.

[0040] The target link combination set refers to a set of optimal link schemes at each time step predicted by the dynamic graph neural network model. The target constellation control task refers to a to-be-planned task selected from the set of to-be-planned constellation control tasks. The constellation control task planning result refers to a final task plan composed of all executable constellation control tasks in the target planning period.

[0041] The technical scheme of the embodiment can optimize constellation control task planning through a dynamic graph neural network, efficiently process large-scale task sets, quickly respond to dynamic changes in space environment and task requirements, solve the shortcomings of existing planning methods in dynamic adaptability and real-time performance, improve the efficiency and effectiveness of constellation control task planning, and improve the operation efficiency and response capability of the constellation system in a complex space environment.

[0042] In an optional manner, the step of obtaining the visible time window of each target antenna of the ground station for any target satellite includes: Based on the overpass prediction information of the any target satellite, at least one visible time window of the any target satellite for each target antenna of the ground station is determined.

[0043] Specifically: ① generate overflight prediction information according to the ephemeris data of any target satellite, and the overflight prediction information contains the orbital parameters and position information of the target satellite in the target planning period. ② Based on the overflight prediction information, calculate the change data of the azimuth and elevation angle of the target satellite relative to each target antenna of the ground station over time. ③ When the elevation angle of the target satellite relative to the target antenna is greater than the preset visible threshold angle, it is determined that the target satellite is visible to the target antenna. ④ Take the time period that continuously satisfies the visible condition as a visible time window of the target satellite, and record the start time point and end time point of the time window. ⑤ Repeat steps ② to ④ until all visible time windows of the target satellite are obtained.

[0044] It should be noted that the way to obtain the visible time window of each target satellite relative to each target antenna of the ground station is the same, and will not be repeated here.

[0045] In the above optional manner, the visible time window is further obtained by using the satellite overflight prediction information, which can accurately determine the communication time range of the satellite and the ground station antenna, provide accurate time basis for subsequent task planning, and enhance the timeliness and feasibility of task planning.

[0046] In an optional manner, the step of constructing a node set corresponding to any time step in the target time step set comprises: Based on all target antennas of the ground station and all target satellites visible at the any time step, a node set corresponding to the any time step is constructed.

[0047] Wherein, the node set corresponding to the tth time step is represented as: ; represents the number of visible target satellites at the tth time step.

[0048] In the above optional manner, the node set is further constructed based on the ground station antenna and the visible satellite, which can clearly present the available resource nodes at each time step, provide accurate input for the dynamic graph neural network model, and help accurately predict the target link combination.

[0049] In an optional manner, the step of constructing an edge set corresponding to any time step in the target time step set comprises: Based on the node set corresponding to the any time step, and combining the target satellites and target antennas having a link relationship at the any time step, an edge set corresponding to the any time step is constructed.

[0050] Wherein, the edge set corresponding to the tth time step is represented as , specifically represented as an edge set of target satellites and target antennas that have a link relationship at the t-th time step, and the connection information of the edge is represented by an adjacency matrix , specifically represented as: when there is a link between the target satellite and the target antenna , the value of the corresponding position is 1, otherwise the value is 0: .

[0051] In the above optional manner, the edge set is further constructed in combination with the node set and the satellite antenna link relationship, accurately describing the connection between nodes, so that the graph neural network model can better capture the interaction information between nodes and improve the rationality of task planning.

[0052] In an optional manner, When the number of target satellites is greater than or equal to the number of target antennas among the target satellites and the target antennas that have a link relationship at the any time step, the target link combination set corresponding to the any time step includes: the antenna-satellite link combination with the highest link probability corresponding to each target antenna that has a link relationship at the any time step.

[0053] When the number of target satellites is less than the number of target antennas among the target satellites and the target antennas that have a link relationship at the any time step, the target link combination set corresponding to the any time step includes: the antenna-satellite link combination with the highest link probability corresponding to each target satellite that has a link relationship at the any time step.

[0054] Wherein, the specific process is: ① for any time step in the target time step set, extract all target satellites and target antennas that have a link relationship at the time step, and count the number of target satellites and the number of target antennas . ② Compare the size relationship of the number of target satellites and the number of target antennas that have a link relationship at the time step. ③ If : for each target antenna that has a link relationship at the time step, select the target satellite with the highest link probability from all possible target satellites that link to the target antenna output by the dynamic graph neural network model. The combination formed by each target antenna and the target satellite with the highest link probability corresponding thereto is taken as the antenna-satellite link combination corresponding to the target antenna. All such combinations constitute the target link combination set corresponding to the time step. If : For each target satellite existing a link relationship at the time step, from all possible target antennas outputted by the dynamic graph neural network model which link to the target satellite, select the target antenna with the highest link probability value. Take the combination of each target satellite and its corresponding target antenna with the highest link probability as the antenna-satellite link combination corresponding to the target satellite. All such combinations constitute the target link combination set corresponding to the time step.

[0055] It should be noted that the way to obtain the target link combination set corresponding to each time step is the same, which is not repeated here.

[0056] In the above optional manner, the target link combination is further determined according to the number relationship between the satellite and the antenna, which can effectively allocate communication resources, select the optimal link scheme under different resource proportion conditions, and improve the resource utilization rate and the task execution success rate.

[0057] In an optional manner, the step of determining a plurality of target constellation TT&C tasks from the set of to-be-planned constellation TT&C tasks according to all target link combination sets comprises: For each to-be-planned constellation TT&C task in the set of to-be-planned constellation TT&C tasks, determine the to-be-planned constellation TT&C task that meets the preset condition as a target constellation TT&C task.

[0058] Wherein, the preset condition is: there is a target time window containing the to-be-controlled time window of the to-be-planned constellation TT&C task, and the preset link combination corresponding to the to-be-planned constellation TT&C task is the same as the target link combination corresponding to the start and end time of the target time window.

[0059] Specifically: ① For any to-be-planned constellation TT&C task in the to-be-planned constellation TT&C task set, the to-be-measured time window required by the to-be-planned constellation TT&C task is extracted. ② A target time window is found in the visible time window set, which needs to completely contain the to-be-measured time window corresponding to the any to-be-planned constellation TT&C task, that is, the start time point and the end time point of the to-be-measured time window are located between the start time point and the end time point of the target time window (which can overlap). ③ The target link combination of the target time step corresponding to the start time point of the target time window and the target link combination of the target time step corresponding to the end time point of the target time window are extracted from the target link combination set. ④ The preset link combination corresponding to the any to-be-planned constellation TT&C task is obtained, which includes the fixed link of the preset target satellite and the target antenna. ⑤ It is judged whether the following two conditions are met at the same time: Condition one: the target link combination corresponding to the start time point of the target time window is the same as the preset link combination of the any to-be-planned constellation TT&C task; Condition two: the target link combination corresponding to the end time point of the target time window is the same as the preset link combination of the any to-be-planned constellation TT&C task. ⑥ If the above two conditions are met at the same time, the any to-be-planned constellation TT&C task is determined as a target constellation TT&C task. ⑦ The processes of ② to ⑥ are repeatedly performed for each to-be-planned constellation TT&C task in the to-be-planned constellation TT&C task set until all to-be-planned constellation TT&C tasks in the to-be-planned constellation TT&C task set are executed, and all target constellation TT&C tasks are obtained.

[0060] In the above optional mode, the target constellation TT&C task is further screened by the preset condition, which can quickly lock the required task, reduce the redundancy and error options in task planning, and improve the accuracy and efficiency of task planning.

[0061] In an optional mode, the constraint condition is: First, the duration constraint condition: the duration of the to-be-measured time window of the target constellation TT&C task is not less than the target duration.

[0062] Specifically: ① The duration of the to-be-measured time window of the target constellation TT&C task is calculated, that is, the difference between the end time point and the start time point. ② It is judged whether the duration is not less than the preset target duration. ③ If it is met, it is determined that the task meets the duration constraint condition; if it is not met, it is determined that the task does not meet the duration constraint condition. It should be noted that the value of the target duration is set according to the actual situation, which is not limited here.

[0063] Second, the visibility constraint condition: the to-be-measured time window of the target constellation TT&C task belongs to any one of the visible time window set.

[0064] Specifically: ①search whether there is a visible time window in the set of visible time windows. ②Judge whether the to-be-controlled time window of the target constellation control task belongs to the time range of the retrieved visible time window, that is, the starting time point of the to-be-controlled time window should be greater than or equal to the starting time point of the visible time window, and the ending time point of the to-be-controlled time window should be less than or equal to the ending time point of the visible time window. ③If it is satisfied, it is determined that the task meets the visibility constraint condition; if it is not satisfied, it is determined that the task does not meet the visibility constraint condition.

[0065] Third, priority constraint condition: when the to-be-controlled time window of the target constellation control task (defined as: the current target constellation control task) has an overlapping time window, the target constellation control task is the highest priority task in the preset priority group.

[0066] Among them, all target constellation control tasks in the overlapping task group have the same time window, and the target satellites and / or target antennas in the preset link combination corresponding to each target constellation control task in the overlapping task group are the same.

[0067] Specifically: ①Identify all other target constellation control tasks that have overlapping (time window intersection) with the to-be-controlled time window of the current target constellation control task. ②According to the same target satellite and / or target antenna in the preset link combination, the tasks with overlapping time windows are divided into the same overlapping task group. ③Get the preset priority of all tasks in the overlapping task group, and determine whether the current target constellation control task is the highest priority task in the overlapping task group. ④If the current target constellation control task is the highest priority, it is determined that the current target constellation control task meets the priority constraint condition; if the current target constellation control task is not the highest priority, it is determined that the current target constellation control task does not meet the priority constraint condition.

[0068] Fourth, link combination constraint condition: the preset link combination of the target constellation control task is any one of all fixed link combinations.

[0069] Specifically: ①Get the preset link combination of the target constellation control task. ②Judge whether the preset link combination is one of all fixed link combinations allowed by the constellation system, that is, whether the combination meets the allowed link rules or configuration between satellites and antennas defined by the constellation system. ③If it is satisfied, it is determined that the task meets the link combination constraint condition; if it is not satisfied, it is determined that the task does not meet the link combination constraint condition.

[0070] Finally, each target constellation control task that simultaneously meets the time length constraint condition, the visibility constraint condition, the priority constraint condition and the link combination constraint condition is determined as an executable constellation control task.

[0071] In the above optional methods, multi-dimensional constraints are further clarified to ensure the feasibility and rationality of the task, guarantee key factors such as task duration and priority, avoid resource conflicts and task overlaps, and make task planning more scientific and effective.

[0072] In this embodiment, the step of using the dynamic graph neural network model to generate a target link combination set corresponding to each time step in the target time step set includes: ① According to the target time step set Divide the constellation system with dynamically changing network topology into a series of static graph snapshots },in is the number of elements in the time step set, and each snapshot is considered as a graph at the split point time step t , A node set consisting of all ground station antennas and visible target satellites at time t (the tth time step) , Indicates the number of visible target satellites at time t (t-th time step); Represents the edge set, which represents the edge set of the link relationship between the satellite and the ground station antenna at time t (the tth time step). The edge connection information is represented by the adjacency matrix Indicates that the calculation formula is as follows: when the target satellite and target antenna If there is a link between them, the value of the corresponding position is 1, otherwise the value is 0: .

[0073] ② Node2Vec is used to map the nodes in the graph to a low-dimensional vector space. Node2Vec generates a graph node sequence through random walks, and then uses skip-gram modeling to obtain the feature representation H of the graph nodes. The probability formula for walk calculation is: ;in, represents the probability of the previous node c swimming to the next node d, represents the shortest distance from node c to node d, Indicates that the next step of the current node is to return to the node c of the previous step. Indicates that the next step of the current node is to return to the neighbor node of the previous step node c, Indicates that the next step of the current node is to walk to a neighboring node farther away from the node c in the previous step. p is used to control the probability of returning to the previous node during the walk. The larger the p is, the more likely the model is to visit farther nodes. q is used to control the direction of the random walk. The larger the value of q is, the more likely the walk is to search deeply, and the smaller the value of q is, the more likely the walk is to search broadly (e.g. Figure 3 shown).

[0074] ③Time series graph convolution layer processing: Step one, define the history window of node c at time t as , aggregate the history states of the neighborhood in the last time steps: ; where is the attention weight, W is the shared weight matrix, is the time interval; Step two, process the state sequence of the node in the time dimension by a one-dimensional convolution kernel K to capture the temporal dependence: .

[0075] (4) Node state gated update: Step one, introduce update gate and reset gate to control the retention and forgetting of historical information: , ; Step two, candidate state generation: ; Step three, state update: ; where ⊙ represents element-wise multiplication, and σ is the Sigmoid function.

[0076] (5) Time information encoding fusion: Step one, convert the timestamp t into an embedding vector using sinusoidal encoding: ; where d is the embedding dimension.

[0077] Step two, relative time encoding: calculate the event interval , and convert it to features by MLP; Step three, concatenate or weighted fusion of time encoding and node state: .

[0078] (6) Link prediction: Step one, extract the states of the target node pair (c, d) at time t and ; Step two, calculate the interaction score: ; Step three, convert to link probability by Sigmoid function: , finally get the link probability of each antenna-satellite link combination corresponding to each time step.

[0079] For the training process of the dynamic graph neural network model, it needs to be noted that: Step one, obtain the dataset by simulating the ground station antenna and satellite using STK software, and get the ground station antenna's visible time window set for each satellite in a period of time , a certain number of target nodes i and target antennas j are randomly selected at each time step to establish a node set and an edge set , and the establishment and disconnection of the simulation TT&C task are simulated, the node set and the edge set are input into the model to obtain vector representation data of the nodes, and the vector representation data is input into the model to obtain the link probability of the nodes. The first 90% of the data in the data set is used as a training set, and the remaining 10% is used as a test set. A negative sampling method can be used to randomly select a part of non-link data as negative sample data, which is used together with the positive sample data for training and testing. The timestamp and node feature H of each time step are used as input, and the model updates the model parameters through back propagation according to the latest graph structure and node behavior after processing each time step, so as to ensure that the model adapts to dynamic changes.

[0080] Step two, select a cross-entropy loss function to measure the difference between the model prediction result and the true label. Select the Adam optimizer to update the model parameters to minimize the loss function. The model is evaluated by the AUC index, and the process is repeated until the model converges or reaches the preset training number of rounds. The trained dynamic graph neural network model, including the structure, parameter weight, and optimizer parameters of the model, can be used for actual TT&C task planning. Step three, use the trained model to plan the constellation TT&C task. The TT&C requirements for a period of time are input into the model, and whether the satellite and the ground station antenna are linked at the next time step is predicted by the model. The antenna-satellite link combination with the highest probability ranking is selected.

[0081] Figure 4 An embodiment of a constellation TT&C task planning device 200 based on a dynamic graph neural network provided by the application is shown in the structural schematic diagram. As Figure 4 shown, the constellation TT&C task planning device 200 based on the dynamic graph neural network includes a determination module 210, a construction module 220, a prediction module 230, and a planning module 240. The determination module 210 is configured to obtain a set of to-be-planned constellation TT&C tasks in a target planning period from user TT&C requirements, and determine target satellites involved in the set of to-be-planned constellation TT&C tasks. The construction module 220 is configured to obtain a visible time window of each target antenna of a ground station for each target satellite, to form a set of visible time windows, and arrange a starting time point and an ending time point corresponding to each visible time window in the set of visible time windows in time sequence to obtain a set of target time steps. The prediction module 230 is configured to: based on the node set and the edge set corresponding to each time step of the target time step set, and by using the trained dynamic graph neural network model, obtain a target link combination set corresponding to each time step in the target time step set. The planning module 240 is configured to: according to all target link combination sets, determine a plurality of target constellation TT&C tasks from the set of to-be-planned constellation TT&C tasks, and determine a target constellation TT&C task meeting a constraint condition as an executable constellation TT&C task to form a constellation TT&C task planning result of the target planning period.

[0082] In an optional manner, the construction module 220 is specifically configured to: Based on the overflight prediction information of any target satellite, determine at least one visible time window of each target antenna of the ground station for the any target satellite.

[0083] In an optional manner, the prediction module 230 is specifically configured to: Based on all target antennas of the ground station and all target satellites visible at the any time step, construct a node set corresponding to the any time step.

[0084] In an optional manner, the prediction module 230 is specifically configured to: Based on the node set corresponding to the any time step, and in combination with target satellites and target antennas having a link relationship at the any time step, construct an edge set corresponding to the any time step.

[0085] In an optional manner, When the number of target satellites is greater than or equal to the number of target antennas among target satellites and target antennas having a link relationship at the any time step, the target link combination set corresponding to the any time step includes: an antenna-satellite link combination having the highest link probability corresponding to each target antenna having a link relationship at the any time step; When the number of target satellites is less than the number of target antennas among target satellites and target antennas having a link relationship at the any time step, the target link combination set corresponding to the any time step includes: an antenna-satellite link combination having the highest link probability corresponding to each target satellite having a link relationship at the any time step.

[0086] In an optional manner, the planning module 240 is specifically configured to: For each to-be-planned constellation TT&C task in the to-be-planned constellation TT&C task set, a to-be-planned constellation TT&C task satisfying a preset condition is determined as a target constellation TT&C task; wherein the preset condition is that there is a target time window containing a to-be-TTCed time window of the to-be-planned constellation TT&C task, and a preset link combination corresponding to the to-be-planned constellation TT&C task is the same as a target link combination corresponding to start and end time of the target time window.

[0087] In an optional manner, the constraint condition is: A time length of the to-be-TTCed time window of the target constellation TT&C task is not less than a target time length; The to-be-TTCed time window of the target constellation TT&C task belongs to any one of the visible time window set; When the to-be-TTCed time window of the target constellation TT&C task has an overlapping time window, the target constellation TT&C task is a task with the highest preset priority in an overlapping task group to which the target constellation TT&C task belongs; wherein all target constellation TT&C tasks in the overlapping task group have the same time window, and target satellites and / or target antennas in preset link combinations corresponding to each target constellation TT&C task in the overlapping task group are the same; The preset link combination of the target constellation TT&C task is any one of all fixed link combinations.

[0088] It should be noted that the beneficial effects of the constellation TT&C task planning device 200 based on the dynamic graph neural network provided in the above embodiments are the same as those of the constellation TT&C task planning method based on the dynamic graph neural network, and will not be repeated here. In addition, when the device provided in the above embodiments implements its functions, only the division of the above functional modules is exemplified, and in actual application, the above functions can be completed by different functional modules according to needs, that is, the device is divided into different functional modules according to actual conditions to complete all or part of the above described functions. In addition, the device and method embodiments provided in the above embodiments belong to the same concept, and the specific implementation process is shown in the method embodiments, which will not be repeated here.

[0089] Among them, the constellation TT&C task planning device 200 based on the dynamic graph neural network of the application can be a computer program (including program code) running in a computer device, for example, the constellation TT&C task planning device 200 based on the dynamic graph neural network of the application is an application software, which can be used to execute corresponding steps in the constellation TT&C task planning method based on the dynamic graph neural network of the application.

[0090] In some embodiments, the constellation measurement and control task planning device 200 based on dynamic graph neural network of the present invention can be implemented by a combination of software and hardware. As an example, the constellation measurement and control task planning device based on dynamic graph neural network of the present invention can be a processor in the form of a hardware decoding processor, which is programmed to execute the constellation measurement and control task planning method based on dynamic graph neural network of the present invention. For example, the processor in the form of a hardware decoding processor can adopt one or more application-specific integrated circuits (ASICs), DSPs, programmable logic devices (PLDs), complex programmable logic devices (CPLDs), field-programmable gate arrays (FPGAs) or other electronic components.

[0091] The modules described in the embodiments of the present invention may be implemented in software or hardware, and the name of a module does not necessarily limit the module itself.

[0092] An electronic device according to an embodiment of the present invention includes a memory, a processor, and a computer program stored in the memory and runnable on the processor. When the processor executes the computer program, any of the above-mentioned constellation measurement and control task planning methods based on dynamic graph neural networks is implemented. That is to say, an electronic device according to an embodiment of the present invention may include but is not limited to: a processor and a memory; the memory is used to store the computer program; the processor is used to execute the constellation measurement and control task planning method based on dynamic graph neural networks shown in any embodiment of the present invention by calling the computer program.

[0093] In an alternative embodiment, an electronic device is provided, such as Figure 5 As shown, Figure 5 The electronic device 4000 shown includes: a processor 4001 and a memory 4003. The processor 4001 and the memory 4003 are connected, for example, via a bus 4002. Optionally, the electronic device 4000 may further include a transceiver 4004, which can be used for data exchange between the electronic device and other electronic devices, such as data transmission and / or data reception. It should be noted that in actual applications, the number of transceivers 4004 is not limited to one, and the structure of the electronic device 4000 does not constitute a limitation on the embodiments of the present invention.

[0094] The processor 4001 can be a CPU (Central Processing Unit), a general-purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array) or other programmable logic device, transistor logic device, hardware component, or any combination thereof. It can implement or execute various exemplary logical blocks, modules and circuits described in connection with the present disclosure. The processor 4001 can also be a combination of computing functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, and the like.

[0095] The bus 4002 can include a path for transmitting information between the above-mentioned components. The bus 4002 can be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus, or the like. The bus 4002 can be divided into an address bus, a data bus, a control bus, and the like. For convenience of representation, Figure 5 The bus 4002 is represented by only one thick line, but it does not mean that there is only one bus or only one type of bus.

[0096] The memory 4003 can be a ROM (Read Only Memory) or other type of static storage device that can store static information and instructions, a RAM (Random Access Memory) or other type of dynamic storage device that can store information and instructions, an EEPROM (Electrically Erasable Programmable Read Only Memory), a CD-ROM (Compact Disc Read Only Memory) or other optical disk storage, a magnetic disk storage medium or other magnetic storage device, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and that can be accessed by a computer, but is not limited thereto.

[0097] The memory 4003 is configured to store application code (computer program) for implementing the scheme of the present application, and the processor 4001 is configured to control the execution. The processor 4001 is configured to execute the application code stored in the memory 4003 to implement the content shown in the foregoing method embodiments.

[0098] The electronic device can also be a terminal device, and the terminal device can be any terminal device that can install an application and access a webpage through the application, including at least one of a smartphone, a tablet computer, a notebook computer, a desktop computer, a smart speaker, a smart watch, a smart television, and a smart vehicle device.

[0099] It should be noted that, Figure 5 The electronic device shown is only an example and should not limit the functions and use range of the embodiments of the present application.

[0100] The computer readable storage medium of the embodiment of the present application, the computer readable storage medium has a computer program stored thereon, and the computer program is executed by the processor to implement any one of the above-mentioned constellation TT&C task planning methods based on dynamic graph neural network.

[0101] Optionally, the computer readable storage medium can be a read-only memory (Read-Only Memory, ROM), a random access memory (Random Access Memory, RAM), a compact disc read-only memory (Compact Disc Read-Only Memory, CD-ROM), a magnetic tape, a floppy disk, and an optical data storage device, etc.

[0102] In the exemplary embodiments, a computer program product or computer program is also provided, which includes computer instructions stored in a computer readable storage medium. The processor of the electronic device reads the computer instructions from the computer readable storage medium, and the processor executes the computer instructions to enable the electronic device to perform the above-mentioned constellation TT&C task planning method based on dynamic graph neural network.

[0103] Computer program code for carrying out operations of the present application can be written in any combination of one or more programming languages, including an object oriented programming language such as Java, Smalltalk, C++ or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code can execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the latter scenario, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection can be made to an external computer (for example, through the Internet using an Internet Service Provider).

[0104] It should be understood that the flowchart and block diagrams in the drawings illustrate the architecture, functionality, and operation of possible implementations of various embodiments of the present application. In this regard, each block in the flowchart and block diagrams can represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions noted in the block can occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently or the blocks can sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and / or flowchart illustrations, and combinations thereof, can be implemented by special purpose hardware-based systems that perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.

[0105] The computer readable storage medium of embodiments of the present application can be, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, or any suitable combination of the foregoing. More specific examples of the computer readable storage medium can include, but are not limited to, an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the present application, the computer readable storage medium can be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device.

[0106] The computer readable storage medium described above bears one or more programs, when the one or more programs are executed by the electronic device, the electronic device executes the method shown in the above embodiment.

[0107] The above description is merely that of the preferred embodiments of the application and of the principles thereof. It is to be understood that the disclosed scope of the application is not limited to the specific combinations of technical features described above, but also encompasses other technical solutions formed by any combinations of the technical features described above or their equivalent features without departing from the disclosed concept. For example, the technical solutions formed by mutually replacing the above features with the technical features disclosed in the application (but not limited to) having similar functions.

[0108] It should be noted that the terms "first", "second" and the like in the specification and claims of the application are used to distinguish similar objects, and represent no limitation on the specific order or sequence of the objects. The order of use of similar objects can be interchanged, so that the embodiments of the application described herein can be implemented in an order other than that illustrated or described.

[0109] Those skilled in the art know that the application can be implemented as an apparatus, a method or a computer program product, therefore, the application can be specifically implemented as follows: it can be complete hardware, complete software (including firmware, resident software, microcode, etc.), or a combination of hardware and software, which is generally referred to as "circuit", "module" or "apparatus" in this paper. In addition, in some embodiments, the application can also be implemented as a computer program product in one or more computer readable media, which contains computer readable program code.

[0110] Although the embodiments of the application have been shown and described above, it should be understood that the above embodiments are exemplary and should not be construed as limiting the application, and those skilled in the art can make changes, modifications, replacements and variations to the above embodiments within the scope of the application.

Claims

1. A constellation measurement and control task planning method based on dynamic graph neural network, characterized in that: include: Obtaining a set of to-be-planned constellation measurement and control tasks within a target planning period from user measurement and control requirements, and determining target satellites involved in the set of to-be-planned constellation measurement and control tasks; Obtaining the visible time window of each target satellite to each target antenna of the ground station to form a visible time window set, and arranging the start time point and the end time point corresponding to each visible time window in the visible time window set in time sequence to obtain a target time step set; Based on the node set and edge set corresponding to each time step in the target time step set, and using the trained dynamic graph neural network model, a target link combination set corresponding to each time step in the target time step set is obtained; According to all target link combination sets, multiple target constellation measurement and control tasks are determined from the set of constellation measurement and control tasks to be planned, and the target constellation measurement and control tasks that meet the constraint conditions are determined as executable constellation measurement and control tasks to form the constellation measurement and control task planning result of the target planning period.

2. The constellation measurement and control task planning method based on dynamic graph neural network according to claim 1 is characterized in that: The steps of obtaining a visible time window of any target satellite to each target antenna of a ground station include: Based on the transit prediction information of the any target satellite, at least one visible time window of the any target satellite to each target antenna of the ground station is determined.

3. The constellation measurement and control task planning method based on dynamic graph neural network according to claim 1 is characterized in that: The step of constructing a node set corresponding to any time step in the target time step set includes: Based on all target antennas of the ground station and all target satellites visible at any time step, a node set corresponding to any time step is constructed.

4. The constellation measurement and control task planning method based on dynamic graph neural network according to claim 3 is characterized in that: The step of constructing an edge set corresponding to any time step in the target time step set includes: Based on the node set corresponding to any time step and in combination with the target satellite and the target antenna that have a link relationship at any time step, an edge set corresponding to any time step is constructed.

5. The constellation measurement and control task planning method based on dynamic graph neural network according to any one of claims 2 to 4, characterized in that: When, among the target satellites and target antennas that have a link relationship at any time step, the number of target satellites is greater than or equal to the number of target antennas, the target link combination set corresponding to the any time step includes: the antenna-satellite link combination with the highest link probability corresponding to each target antenna that has a link relationship at the any time step; When the number of target satellites and target antennas that have a link relationship at any time step is less than the number of target antennas, the target link combination set corresponding to the any time step includes: the antenna-satellite link combination with the highest link probability corresponding to each target satellite that has a link relationship at the any time step.

6. The constellation measurement and control task planning method based on dynamic graph neural network according to claim 5 is characterized in that: The step of determining a plurality of target constellation measurement and control tasks from the set of constellation measurement and control tasks to be planned according to all target link combination sets includes: For each to-be-planned constellation measurement and control task in the to-be-planned constellation measurement and control task set, a to-be-planned constellation measurement and control task that meets a preset condition is determined as a target constellation measurement and control task; wherein the preset condition is: there is a target time window that includes the to-be-planned measurement and control time window of the to-be-planned constellation measurement and control task, and the preset link combination corresponding to the to-be-planned constellation measurement and control task is the same as the target link combination corresponding to the start and end times of the target time window.

7. The constellation measurement and control task planning method based on dynamic graph neural network according to claim 6 is characterized in that: The constraints are: The duration of the target constellation tracking and control time window is not less than the target duration; The time window to be measured and controlled of the target constellation measurement and control task belongs to any visible time window in the visible time window set; When the target constellation measurement and control time windows of the target constellation measurement and control task have overlapping time windows, the target constellation measurement and control task is the task with the highest preset priority in the overlapping task group to which it belongs; wherein all the target constellation measurement and control tasks in the overlapping task group have the same time window and each target constellation measurement and control task in the overlapping task group corresponds to the same target satellite and / or target antenna in the preset link combination; The preset link combination of the target constellation tracking and control mission is any one of all fixed link combinations.

8. A constellation measurement and control task planning device based on dynamic graph neural network, characterized in that: include: Determine module, build module, predict module, and plan module; The determination module is used to: obtain a set of to-be-planned constellation measurement and control tasks within a target planning period from user measurement and control requirements, and determine target satellites involved in the set of to-be-planned constellation measurement and control tasks; The construction module is used to: obtain a visible time window of each target satellite to each target antenna of the ground station to form a visible time window set, and arrange the start time point and the end time point corresponding to each visible time window in the visible time window set in time sequence to obtain a target time step set; The prediction module is used to: obtain a target link combination set corresponding to each time step in the target time step set based on the node set and edge set corresponding to each time step in the target time step set and using a trained dynamic graph neural network model; The planning module is used to: determine multiple target constellation measurement and control tasks from the set of constellation measurement and control tasks to be planned based on all target link combination sets, and determine the target constellation measurement and control tasks that meet the constraint conditions as executable constellation measurement and control tasks to form the constellation measurement and control task planning results of the target planning period.

9. An electronic device, characterized in that: The electronic device includes a processor, the processor is coupled to a memory, and the memory stores at least one computer program. The at least one computer program is loaded and executed by the processor, so that the electronic device implements the constellation measurement and control task planning method based on a dynamic graph neural network as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that At least one computer program is stored in the computer-readable storage medium, and the at least one computer program is loaded and executed by the processor so that the computer-readable storage medium implements the constellation measurement and control task planning method based on dynamic graph neural network as described in any one of claims 1 to 7.

Images