Dynamic scheduling management method and system for space measurement and control mission resources

Abstract

The present invention provides a method and system for dynamic scheduling management of space measurement and control mission resources, relating to the technical field of space measurement and control resource scheduling. The method divides space measurement and control tasks into routine tasks, fast-response tasks, and five types of composite-constrained tasks, and defines task characteristic vectors based on the time sensitivity of the tasks, the intensity of resource requirements, and the coupling degree between tasks; satellite health parameters, equipment operating conditions, and environmental data are collected in real time through satellite-ground IoT terminals to construct a digital twin; tasks are prioritized, resources are matched, and conflict resolution mechanisms are implemented based on the task characteristic vectors, and space measurement and control tasks are globally optimized and scheduled, with cross-domain two-level scheduling being performed during global optimization scheduling; based on bitmap decision matrix calculation, weighted calculation is performed in combination with task matching values ​​and conflict depths, the priorities of conflicting tasks are dynamically adjusted, and conflicts generated during the scheduling process are dynamically resolved. The method can be applied to resource scheduling and optimization of low-orbit satellites and ground stations.

Description

Technical Field

[0001] The present invention relates to the technical field of aerospace measurement and control resource scheduling, and in particular to a method and system for dynamic scheduling management of aerospace measurement and control mission resources. Background Art

[0002] With the advancement of space technology, the number of satellites in low-orbit constellations has increased annually, placing a growing demand on ground-based measurement and control (TT&C) resources. Current aerospace TT&C resource scheduling faces multiple technical bottlenecks, including insufficient dynamic task response, tight resource coupling, difficulties in multi-task coordination, insufficient utilization of historical data, and a lack of cross-domain resource coordination. These issues lead to high latency, resource conflicts, and low scheduling efficiency in aerospace TT&C task scheduling, making it impossible to meet the demands for rapid response and efficient resource allocation. Summary of the Invention

[0003] In view of this, the purpose of the present invention is to propose a dynamic scheduling management method and system for aerospace measurement and control mission resources, aiming to solve the problems of high concurrency, task conflict, resource allocation delay and other problems existing in the prior art of aerospace measurement and control resource scheduling through multi-modal dynamic collaborative scheduling. The present invention can be applied to resource scheduling and optimization of low-orbit satellites and ground stations, especially in scenarios such as high concurrency tasks, rapid response and emergency measurement and control, and has important application value.

[0004] To achieve the above object, the present invention provides the following technical solutions:

[0005] Based on the above objectives, in a first aspect, the present invention provides a method for dynamic scheduling and management of space tracking and control mission resources, comprising the following steps:

[0006] Space tracking and control tasks are divided into routine tasks, fast-response tasks, and five types of complex-constraint tasks. Task feature vectors are defined based on the time sensitivity, resource demand intensity, and inter-task coupling of the tasks.

[0007] Satellite health parameters, equipment operating conditions, and environmental data are collected in real time through satellite-ground IoT terminals to build a digital twin, and the real-time data is fed back to the intelligent scheduling engine.

[0008] The intelligent scheduling engine combines a distributed DQN network and a particle swarm algorithm to prioritize tasks, match resources, and resolve conflicts based on task feature vectors, enabling global optimization and scheduling of space tracking and control tasks.

[0009] Based on the bitmap decision matrix calculation, the task matching value and conflict depth are combined for weighted calculation to dynamically adjust the priority of conflicting tasks and dynamically resolve conflicts generated during the scheduling process.

[0010] During global optimization scheduling, cross-domain two-level scheduling is performed. The first-level scheduling generates candidate resources based on the task visibility window and domain load, and the second-level scheduling generates the final scheduling plan through local conflict resolution, supporting the coordinated scheduling of space-based and ground-based resources.

[0011] As a further solution of the present invention, the task feature vector is defined as T = [τ, μ, σ], where τ is the time sensitivity of the task, which is the difference between the deadline of the task and the available time window; μ is the resource requirement intensity of the task, which indicates the number of resources required for task execution; σ is the coupling degree of the task, which indicates the mutual dependence between tasks.

[0012] As a further solution of the present invention, in the task feature vector:

[0013] Time sensitivity τ is the difference between the duration of the available time window and the remaining execution time of the task;

[0014] The resource requirement intensity μ is the normalized value of the sum of the number of antenna beams and computing units required for task execution;

[0015] The coupling degree σ is obtained by calculating the data dependency matrix between tasks. When there is equipment sharing or time window overlap, the σ value increases by 20%-50%.

[0016] As a further solution of the present invention, the processing of the five types of composite constraint tasks includes:

[0017] Synchronous tracking tasks: Using sliding time window scanning, the system searches for the intersection of available time windows of multiple devices with a preset step size. When the duration of the intersection is greater than or equal to the minimum task requirement, the system automatically locks the optimal start time.

[0018] Multi-station relay missions: Implement a split strategy of prioritizing difficult tasks before easy ones, prioritize the allocation of scarce Ka-band resources, and allow mission arcs to dynamically adjust time windows within a ±15-second range. Generate a network replenishment plan to support mission arc preemption and replacement.

[0019] Cross-domain collaborative tasks: Through a two-level scheduling process, the first-level scheduling generates candidate resources based on the visible window and domain load, and the second-level scheduling generates the final solution through local conflict resolution. In the first-level scheduling stage, at least 15% of the fiber link bandwidth resources are reserved;

[0020] Emergency rescue tasks: A dynamic preemption strategy is adopted to allow tasks with a priority level ≥ 2 to interrupt low-coupling regular tasks;

[0021] Multi-objective joint task: Through the resource time-sharing multiplexing mechanism, the task is split into multiple sub-stages and allocated independent resource pools.

[0022] As a further solution of the present invention, the distributed DQN network adopts a dual-channel feature fusion mechanism, including:

[0023] The spatial channel uses a graph convolutional network to model the topological relationship between space-based and ground-based resources. The nodes include satellite orbit altitude, ground station longitude and latitude, and link attenuation parameters.

[0024] The time series channel uses LSTM to capture the dynamic evolution characteristics of the task queue and updates the hidden state every 60 seconds;

[0025] The dual-channel weights are dynamically balanced through the attention mechanism, and the spatial channel dominant mode (α ≥ 0.7) is activated when the resource conflict rate is > 30%.

[0026] As a further solution of the present invention, based on the bitmap decision matrix calculation, when combining the task matching value and the conflict depth for weighted calculation, the task matching value M j and conflict depth C i The weighted formula is:

[0027]

[0028] Among them, D ij Indicates the scheduling conflict between task i and task j. The larger the value, the higher the conflict degree, which needs to be resolved first. i Indicates the conflict depth of task i, which indicates the degree of conflict between task i and other tasks during the scheduling process. The greater the conflict depth, the more priority the task needs to be scheduled; M j represents the matching value of task j, which indicates the matching degree between task j and resources; T i represents the timeliness of task i; ω1, ω2, ω3 represent the weighted coefficients of conflict depth, task matching degree and timeliness respectively; λ is a dynamic attenuation factor, which represents the impact of task timeliness on scheduling priority. When task timeliness T i When the threshold is reached, λ=1, otherwise, λ=0.5.

[0029] As a further solution of the present invention, when prioritizing tasks, matching resources and resolving conflicts according to task feature vectors, a short message driven mode is adopted to complete resource matching and conflict resolution within 10 seconds, and the emergency task scheduling delay is ≤5 seconds. When sudden tasks are inserted, the resource allocation weight is adjusted by the dynamic attenuation factor λ to give priority to the execution of high-priority tasks.

[0030] As a further solution of the present invention, the coordinated scheduling of space-based and ground-based resources includes:

[0031] Categorize tracking and control resources into two categories: space-based and ground-based, and four status categories: planned, idle, transferable, and preemptible, for a total of eight resource combinations. Priority will be given to applying for idle space-based resources and ground-based resources with minimal conflicts.

[0032] Priority migration and cross-domain conflict resolution are performed on conflicting tasks, supporting concurrent scheduling of multiple users and multiple tasks.

[0033] As a further solution of the present invention, the method for dynamic scheduling and management of space measurement and control mission resources further includes establishing a dynamic resource reservation pool mechanism, wherein the dynamic resource reservation pool mechanism includes:

[0034] Mandatory reservation of no less than two space-based Ku-band beam resources for fast-response missions;

[0035] Pre-allocate 15% of the fiber link bandwidth between ground and base stations for cross-domain collaborative tasks;

[0036] Reserve resources by Exponential decay release, where R release is the amount of reserved resources that should be released at the current moment, R reserved is the total amount of initially reserved resources, β is the dynamic recovery coefficient, and the β value is dynamically adjusted in the range of 0.1-0.5 according to the urgency of the task. β = 0.5 for fast-response tasks and β = 0.2 for regular tasks. e is the base of the natural logarithm, t is the current time, and t0 is the start time of resource reservation.

[0037] In a second aspect, the present invention further provides a dynamic scheduling and management system for space measurement and control mission resources, comprising the following components:

[0038] Satellite-ground IoT terminal, used to collect satellite attitude, power supply voltage and RF power data in real time, with a sampling frequency of ≥10Hz;

[0039] A digital twin construction module connects the satellite-ground IoT terminals to build a dynamic mapping system including a satellite orbit prediction model and an equipment health assessment model;

[0040] An intelligent scheduling engine, comprising a distributed DQN network and a particle swarm algorithm hybrid computing unit, configured to execute the aforementioned dynamic scheduling management method for space tracking and control mission resources;

[0041] A cross-domain scheduling coordinator is used to implement a two-level scheduling strategy, including a first-level scheduler based on a resource topology map and a second-level scheduler based on a conflict resolution matrix;

[0042] Performance evaluation module, real-time calculation of task on-time completion rate η time , beam reuse index μ beam and maneuver cost coefficient C move and feeds the evaluation results back to the intelligent scheduling engine.

[0043] As a further solution of the present invention, the intelligent scheduling engine includes:

[0044] a conflict resolution unit configured to mark resource occupancy status in a bitmap and trigger dynamic priority migration when a time window overlap of ≥3 minutes is detected;

[0045] The emergency response unit is configured to complete resource preemption and rescheduling of existing tasks within 5 seconds upon receiving an emergency task with a priority level ≥ 2;

[0046] The fault-tolerant rescheduling unit is configured to run conservative and aggressive scheduling schemes in parallel, and automatically switches to the conservative scheme when the device health deviation is >20%.

[0047] As a further solution of the present invention, the digital twin building module includes:

[0048] The anomaly detection submodule is configured to generate a level 3 alarm when the satellite battery temperature mutation rate exceeds 5°C / min;

[0049] Resource visualization submodule, dynamically displays the orbital distribution map of space-based resources and the heat map of ground-based resource load, with a refresh rate of ≥1 time / second;

[0050] The predictive maintenance submodule predicts the remaining life of the equipment based on the LSTM network and starts the spare parts scheduling process when the predicted life is less than 30 days.

[0051] As a further embodiment of the present invention, the performance evaluation module includes:

[0052] Three-dimensional measurement unit, calculating the on-time completion rate of tasks:

[0053]

[0054] Among them, W ontime is the number of tasks completed on time, with an error of ≤±1 minute; W total is the total number of tasks;

[0055] Calculate the beam reuse index:

[0056]

[0057] in, is the actual usage time of a single beam; N beam is the total number of beams in the system; T total is the total duration of the statistical period;

[0058] Calculate the mobility cost coefficient:

[0059]

[0060] in, is the antenna azimuth adjustment value of the i-th task; is the antenna pitch angle adjustment value of the i-th task; Paz Adjust the cost per azimuth unit; P el Adjusted cost for pitch angle units;

[0061] Mode switching unit, when η time <90% and μ beam When the task rate is >1.2, the batch scheduling ratio of routine tasks is increased to more than 75%;

[0062] Cost Optimization Unit, according to C move The value automatically selects the phased array radar combination solution with the lowest maneuvering cost.

[0063] Compared with the existing technology, the method and system for dynamic scheduling and management of space tracking and control mission resources proposed in the present invention have the following beneficial effects:

[0064] 1. This invention divides tasks into routine tasks, fast-response tasks, and five types of complex-constrained tasks, and constructs a three-dimensional feature vector based on time sensitivity, resource demand intensity, and coupling degree. This allows for precise quantification of task priorities and improves priority sorting efficiency. The intelligent scheduling engine combines the global optimization capabilities of deep reinforcement learning with the rapid convergence characteristics of the particle swarm algorithm to quickly complete complex task scheduling decisions and support high-concurrency task processing.

[0065] 2. The present invention also adopts a two-level conflict resolution mechanism for dynamic conflict resolution and optimized resource utilization. The first-level scheduling generates candidate resources based on the task visibility window and domain load, dynamically calculates the conflict degree through a bitmap decision matrix, and shortens the conflict detection response time. The second-level scheduling uses a local conflict resolution matrix, combined with task matching value and conflict depth weighted calculation, to reduce the resource conflict rate. Furthermore, space-based Ku-band resources are forcibly reserved for fast-response tasks, and fiber bandwidth is pre-allocated for cross-domain collaborative tasks, dynamically releasing idle resources and improving resource utilization.

[0066] 3. In the cross-domain two-level scheduling strategy of the present invention, the first-level scheduling generates candidate resources based on the resource topology map, and supports the four-layer and eight-category dynamic division of space-based / ground-based resources; the second-level scheduling realizes multi-domain resource collaboration through the conflict resolution matrix, improves the success rate of cross-domain task scheduling, adopts a short message drive mode, shortens the emergency task scheduling delay, adjusts the weight through the dynamic attenuation factor when inserting sudden tasks, and improves the success rate of high-priority task preemption.

[0067] In summary, the present invention systematically solves core problems in the field of aerospace measurement and control, such as high-concurrency task conflicts, resource allocation delays, and cross-domain collaboration difficulties, through task feature modeling, multi-algorithm fusion scheduling, dynamic resource management, and cross-domain collaboration mechanisms. It significantly improves the real-time performance, reliability, and resource utilization efficiency of the measurement and control system, and provides key technical support for future large-scale constellation management and emergency measurement and control. It supports dynamic matching of resources between large-scale low-orbit satellites (such as Starlink and remote sensing constellations) and ground stations, improves measurement and control coverage, and is also suitable for scenarios such as satellite failure rescue and space threat avoidance.

[0068] These and other aspects of the present application will be more clearly understood in the following description of the embodiments. It should be understood that the above general description and the following detailed description are merely exemplary and explanatory and cannot limit the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or related technologies, the following briefly introduces the drawings required for the exemplary embodiments or related technical descriptions. The drawings are used to provide a further understanding of the present invention and constitute part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the drawings:

[0070] Figure 1 The present invention provides a flow chart of a method for dynamic scheduling and management of space tracking and control mission resources according to an embodiment of the present invention.

[0071] Figure 2 The present invention provides a flowchart of a distributed DQN network using a dual-channel feature fusion mechanism in a method for dynamic scheduling and management of space measurement and control mission resources according to an embodiment of the present invention.

[0072] Figure 3 This is a structural block diagram of a dynamic scheduling and management system for space measurement and control mission resources according to an embodiment of the present invention. DETAILED DESCRIPTION

[0073] Below, the present application is further described in conjunction with the accompanying drawings and specific implementation methods. It should be noted that, under the premise of no conflict, the various embodiments or technical features described below can be arbitrarily combined to form new embodiments.

[0074] To make the purpose, technical solutions and advantages of the present invention more clearly understood, the following is a further detailed description of the embodiments of the present invention in conjunction with specific embodiments and with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are only used to explain the present application and are not intended to limit the present application.

[0075] It should be noted that all expressions using "first" and "second" in the embodiments of the present invention are intended to distinguish two non-identical entities or non-identical parameters with the same name. Therefore, "first" and "second" are used for convenience of expression only and should not be understood as limitations on the embodiments of the present invention. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, other steps or units inherent to a process, method, system, product, or device that includes a series of steps or units.

[0076] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0077] The flowcharts shown in the accompanying drawings are for illustrative purposes only and do not necessarily include all contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps may be decomposed, combined, or partially merged, so the actual execution order may vary depending on the actual situation.

[0078] The following describes some embodiments of the present application in detail with reference to the accompanying drawings. In the absence of conflict, the following embodiments and features therein may be combined with each other.

[0079] In response to the problems of insufficient dynamic task response, strong resource coupling, difficulty in multi-task collaboration, insufficient utilization of historical data and lack of cross-domain resource collaboration in the existing aerospace measurement and control resource scheduling, the present invention proposes a dynamic scheduling and management method for aerospace measurement and control task resources, aiming to solve the problems of high concurrency, task conflict, resource allocation delay and other problems in the existing aerospace measurement and control resource scheduling through multi-modal dynamic collaborative scheduling. The present invention can be applied to the resource scheduling and optimization of low-orbit satellites and ground stations, and has important application value, especially in scenarios such as high concurrency tasks, rapid response and emergency measurement and control.

[0080] See also Figure 1 As shown, an embodiment of the present invention provides a method for dynamic scheduling and management of space measurement and control mission resources, the method comprising the following steps:

[0081] Step S10: Divide the space tracking and control tasks into routine tasks, fast-response tasks, and five types of complex-constrained tasks, and define task feature vectors based on the time sensitivity, resource demand intensity, and inter-task coupling of the tasks.

[0082] In this step, the defined task feature vector is T = [τ, μ, σ], where τ is the time sensitivity of the task, which is the difference between the task deadline and the available time window; μ is the resource demand intensity of the task, which indicates the number of resources required for task execution; and σ is the coupling degree of the task, which indicates the interdependence between tasks.

[0083] Among them, in the task feature vector:

[0084] Time sensitivity τ is the difference between the duration of the available time window and the remaining execution time of the task, that is: τ = T window -(T deadline -T current ), where T window is the total duration of the available time window for the task, T deadline is the absolute deadline of the task, T current is the current time. For example, if the available time window for a task is 30 minutes, the current time is 10:00, and the task deadline is 10:05, then the time sensitivity τ = 30 - (10:05 - 10:00) = 30 - 5 = 25 minutes. The smaller the time sensitivity τ, the more urgent the remaining available time for the task, and the more priority it needs to be scheduled.

[0085] The resource demand intensity μ is the normalized value of the sum of the number of antenna beams and computing units required for task execution. For example, if the task requires two Ka-band antennas and three computing units, and the resource pool has a total of 10 antennas and 10 computing units, then the resource demand intensity μ = (2 + 3) / 20 = 0.25. The higher the resource demand intensity μ, the greater the resource usage, and resource overload must be avoided.

[0086] The coupling degree σ is calculated by the data dependency matrix between tasks. If there is equipment sharing or time window overlap between tasks, the coupling degree σ value increases by 20%-50%. For example, if tasks A and B share the same ground station, then σ A and σ B Each increases by 30%. The higher the coupling degree σ, the stronger the inter-task dependency and the more coordinated scheduling is required. Among them, the inter-task data dependency matrix is ​​a square matrix D=d used to quantify the degree of coupling between tasks. ij ], where the matrix element d ij Indicates the dependency strength between task i and task j, with a value range of [0,1]. If task i and task j share the same device (such as an antenna), the matrix element d ij Increase by 30%; if the task time windows overlap by more than 10%, the matrix element d ij Increase by 20%; other situations (such as data flow dependency) are weighted according to actual needs.

[0087] For example, assuming that mission A and mission B share the same ground station and their time windows overlap by 15%, then:

[0088] d AB =d BA =0.3 (equipment sharing) + 0.2 (time window overlap) = 0.5; this matrix is ​​used by the scheduling engine to determine resource conflicts between tasks, and tasks with high coupling need to be coordinated and scheduled.

[0089] Among these tasks, routine tasks include daily satellite orbit maintenance and routine data transmission, with a fixed time window. For example, if the time window is fixed from 2:00 PM to 2:30 PM daily, the resource requirement intensity μ = 0.3, requiring one S-band antenna. For fast-response tasks, such as sudden disaster monitoring, such as earthquakes and forest fires, the time sensitivity τ = 5 minutes (with tight deadlines) and μ = 0.8.

[0090] In this embodiment, the processing of the five types of compound constraint tasks includes:

[0091] Synchronous tracking tasks: Use sliding time window scanning to search for the intersection of available time windows of multiple devices with a preset step size (such as 1 second). When the intersection duration is ≥ the minimum requirement of the task (such as 10 seconds), the optimal start time is automatically locked.

[0092] Multi-station relay missions: Implement a split strategy of prioritizing difficult tasks before easy ones, prioritize the allocation of scarce resources (such as the Ka band), allow mission arcs to dynamically adjust time windows within a ±15-second range, and generate network replenishment plans to support mission arc preemption and replacement.

[0093] Cross-domain collaborative tasks: Through a two-level scheduling process, the first-level scheduling generates candidate resources based on the visible window and domain load, and the second-level scheduling generates the final solution through local conflict resolution. In the first-level scheduling stage, at least 15% of the fiber link bandwidth resources are reserved;

[0094] Emergency rescue tasks: A dynamic preemption strategy is adopted to allow tasks with a priority level ≥ 2 to interrupt low-coupling regular tasks;

[0095] Multi-target joint mission: Through the resource time-sharing multiplexing mechanism, the task is split into multiple sub-stages (such as tracking, data transmission, and processing) and independent resource pools (such as antennas in different frequency bands) are allocated.

[0096] Step S20: Collect satellite health parameters, equipment operating conditions, and environmental data in real time through the satellite-ground IoT terminal, build a digital twin, and feed the real-time data back to the intelligent scheduling engine.

[0097] In this step, satellite health parameters, equipment operating conditions and environmental data are collected in real time through the satellite-ground IoT terminal. The satellite health parameters include: battery voltage (sampling frequency 10Hz), gyroscope deviation (accuracy 0.01° / s), and solar panel temperature (-50℃~+80℃).

[0098] Equipment operating conditions include: antenna azimuth (0° to 360°), transmission power (10W to 100W), and computing unit load rate (0% to 100%).

[0099] Environmental data include: space radiation dose rate (μGy / h), plasma density (10 6 / m 3 ).

[0100] In this embodiment, when building a digital twin and predicting satellite orbits, the SGP4 / SDP4 model is used. Two rows of orbital parameters (TLE) are input to calculate the orbit in real time for the next 6 hours with an error of <1km. For example, after the TLE parameters of a low-orbit satellite (orbital altitude 550km) are updated, its future time window for passing through the Beijing ground station is predicted. Among them, SGP4 (Simplified General Perturbations 4) is an orbit prediction model for low-Earth orbit (LEO) satellites; SDP4 (Simplified Deep Space Perturbations 4) is an extended model for geosynchronous orbit (GEO) or deep space satellites. The SGP4 / SDP4 model inputs two rows of orbital parameters (TLE) and outputs the orbital position (longitude, latitude, altitude) and velocity of the satellite at any time, with a prediction error of <1km. For example, the TLE of the International Space Station (ISS) is calculated using the SGP4 model, and its time window for passing through Beijing in the next 6 hours can be predicted.

[0101] Among them, the TLE format of the two-line track parameters (TLE) is:

[0102] Line 1: Satellite number, orbit inclination (°), right ascension of ascending node (°), eccentricity (×10 7 ), perigee angle (°), mean anomaly angle (°), number of orbits around the earth per day, etc.

[0103] Line 2: Satellite number, orbital period (minutes), semi-major axis (multiples of the Earth's radius), atmospheric drag coefficient, etc.

[0104] In this embodiment, device health assessment uses an LSTM network to predict battery life, taking voltage, temperature, and the number of charge and discharge cycles as inputs, and outputting the remaining life (with an accuracy of ±3 days). Antenna health is assessed by calculating transmission efficiency (actual power / nominal power). If it is <85%, the device is marked as sub-healthy, triggering a maintenance alert. A satellite-ground station topology map is updated every second, with node attributes including satellite orbit altitude (e.g., 550km), ground station latitude and longitude (e.g., Beijing 116°E, 40°N), and link attenuation (-120dB).

[0105] For example, when the battery temperature of a satellite suddenly rises from 25°C to 40°C (sudden change rate 15°C / min), the digital twin triggers a level 3 alarm, predicts that the remaining life will drop from 30 days to 10 days, and initiates the spare parts scheduling process.

[0106] Step S30: The intelligent scheduling engine combines the distributed DQN network and the particle swarm algorithm to prioritize tasks, match resources, and resolve conflicts based on task feature vectors, and perform global optimization scheduling for space tracking and control tasks.

[0107] In this step, see Figure 2 As shown, the distributed DQN network adopts a dual-channel feature fusion mechanism, including the following steps:

[0108] Step S301: The spatial channel uses a graph convolutional network to model the topological relationship between space-based and ground-based resources. The nodes include the satellite orbit altitude, the longitude and latitude of the ground station, and the link attenuation parameters.

[0109] The input for the spatial channel is a space-based-ground resource topology graph (nodes = satellites / ground stations, edges = link parameters). This graph represents the connectivity between space-based (satellites) and ground-based (ground stations) resources. Nodes include: satellites: orbit altitude, available frequency band (S / Ka / Ku), and link attenuation (e.g., -120 dB); ground stations: latitude and longitude, antenna elevation angle (visible when ≥5°), and maximum tracking rate (° / s). Edges include: the visible time window between the satellite and ground station, link quality (signal-to-noise ratio), and data transmission rate (Mbps). For example, the visible time window between a low-orbit satellite (550 km) and a ground station in Beijing (116°E, 40°N) is 2:00 PM to 2:30 PM, with a link attenuation of -110 dB and a data rate of 50 Mbps.

[0110] In this embodiment, the graph convolutional network (GCN) is a three-layer convolution that outputs topological relationship weights. The node relationship (such as the link attenuation between the satellite and the ground station) is modeled through a three-layer graph convolutional network (GCN). The input of the timing channel is the dynamic characteristics of the task queue (task arrival time, priority, resource requirements). The hidden layer of the LSTM network is 64 units. The task arrival trend is captured through LSTM (64 hidden layer units) and the status is updated every 60 seconds. The attention mechanism is used to dynamically balance the weights of the two channels. When the resource conflict rate is >30%, the spatial channel weight α = 0.8 and the temporal channel weight β = 0.2, giving priority to the geographical resource distribution.

[0111] Step S302: The time series channel uses LSTM to capture the dynamic evolution characteristics of the task queue and updates the hidden state every 60 seconds;

[0112] Step S303: Dynamically balance the dual-channel weights through the attention mechanism, and activate the spatial channel dominant mode (α≥0.7) when the resource conflict rate is greater than 30%.

[0113] In this embodiment, when prioritizing tasks, matching resources, and resolving conflicts based on task feature vectors, a short message driven mode is adopted to complete resource matching and conflict resolution within 10 seconds, and the emergency task scheduling delay is ≤5 seconds. When sudden tasks are inserted, the resource allocation weight is adjusted through the dynamic attenuation factor λ to prioritize the execution of high-priority tasks.

[0114] When using a particle swarm optimization algorithm to quickly converge to the optimal resource allocation solution, the particle swarm algorithm parameters were set as follows: number of particles = 50, number of iterations = 100, inertia weight = 0.8, and learning factors c1 = c2 = 1.5. When an emergency task is inserted, the particle swarm algorithm can adjust the resource allocation plan within 5 seconds, ensuring that high-priority tasks are successfully preempted. During the scheduling process, task priority scores are calculated based on the eigenvectors to prioritize tasks. The short message-driven mode is used to complete matching within 10 seconds. Emergency tasks are matched with resource delays of ≤ 5 seconds, and the dynamic attenuation factor λ adjusts the weight.

[0115] The present invention divides tasks into routine tasks, fast-response tasks and five types of complex-constrained tasks, and constructs a three-dimensional feature vector based on time sensitivity, resource demand intensity and coupling degree, thereby achieving accurate quantification of task priorities and improving priority sorting efficiency; the intelligent scheduling engine combines the global optimization capabilities of deep reinforcement learning and the rapid convergence characteristics of the particle swarm algorithm to quickly complete complex task scheduling decisions and support high-concurrency task processing.

[0116] Step S40 : Based on the bitmap decision matrix calculation, weighted calculation is performed in combination with the task matching value and the conflict depth, the priority of the conflicting tasks is dynamically adjusted, and the conflicts generated in the scheduling process are dynamically resolved.

[0117] In this step, based on the bitmap decision matrix calculation, when combining the task matching value and conflict depth for weighted calculation, the task matching value M j and conflict depth C i The weighted formula is:

[0118]

[0119] Among them, D ij Indicates the scheduling conflict between task i and task j. The larger the value, the higher the conflict degree, which needs to be resolved first. i Indicates the conflict depth of task i, which indicates the degree of conflict between task i and other tasks during the scheduling process. The greater the conflict depth, the more priority the task needs to be scheduled; M j represents the matching value of task j, which indicates the matching degree between task j and resources; Ti represents the timeliness of task i; ω1, ω2, ω3 represent the weighted coefficients of conflict depth, task matching degree and timeliness respectively; λ is a dynamic attenuation factor, which represents the impact of task timeliness on scheduling priority. When task timeliness T i When the threshold is reached, λ=1, otherwise, λ=0.5.

[0120] For example, when T i <10 minutes, λ=1, otherwise, λ=0.5. i =2,M j =0.9, T i = 5 minutes. Weights ω1 = 0.6, ω2 = 0.3, ω3 = 0.1, then:

[0121]

[0122] If D ij >0.5, triggering priority adjustment, and task i is scheduled in advance.

[0123] Among them, the bitmap decision matrix is ​​a two-dimensional matrix B = [b rt ], indicating the occupancy status of resource r in time slice t; element b rt is 0 or 1, element b rt When it is 0, it means that the resource r time slice t is idle; element b rt When it is 1, it means that resource r is occupied. If task A applies for resource r in the time window [t1, t2], check For example, if a Ka-band antenna is occupied by Task B from 2:00 PM to 2:15 PM, a conflict will occur when Task A requests the same resource, and the conflict must be resolved by adjusting the priority.

[0124] Step S50: During global optimization scheduling, cross-domain two-level scheduling is performed. The first-level scheduling generates candidate resources based on the task visibility window and domain load, and the second-level scheduling generates the final scheduling plan through local conflict resolution, supporting the coordinated scheduling of space-based and ground-based resources.

[0125] In this step, the coordinated scheduling of space-based and ground-based resources includes:

[0126] Categorize tracking and control resources into two categories: space-based and ground-based, and four status categories: planned, idle, transferable, and preemptible, for a total of eight resource combinations. Priority will be given to applying for idle space-based resources and ground-based resources with minimal conflicts.

[0127] Priority migration and cross-domain conflict resolution are performed on conflicting tasks, supporting concurrent scheduling of multiple users and multiple tasks.

[0128] For example, during the first-level scheduling, global resource pre-selection is performed. When calculating the visibility window, the visibility time window for the next 2 hours is calculated based on the satellite orbit and the geographical location of the ground station (for example, the Beijing station is visible from 10:00 to 10:30). When balancing domain load, domains with a load < 70% are selected (for example, the space-based domain has a load of 60% and the ground-based domain has a load of 75%), and space-based resources are allocated first. For example, the candidate resources for a low-orbit satellite are a space-based relay satellite (load 55%) and a Sanya station (load 80%). The relay satellite is selected for the first-level scheduling.

[0129] During secondary scheduling, local conflict resolution is performed. When resolving the conflict matrix, the Hungarian algorithm is used to match tasks and resources to minimize the overall conflict level. During time window fine-tuning, task time windows are allowed to slide within ±15 seconds to resolve resource overlap. For example, Task A was originally scheduled to use Antenna 1 from 10:00 AM to 10:15 AM, conflicting with Task B. This was fine-tuned to 10:00 AM to 10:14:45 AM, eliminating the conflict.

[0130] During space-based and ground-based collaboration, space-based resources are marked as "preemptible" and ground-based resources are marked as "transferable", supporting cross-domain priority migration. For example, after a space-based relay satellite resource is preempted by an emergency task, the original task is automatically migrated to a ground-based station (such as the Kashgar station). Resources in space-based and ground-based collaboration are classified into space-based resources and ground-based resources. Space-based resources include relay satellites (such as Tianlian) and navigation satellites (such as Beidou); ground-based resources include ground stations, fiber optic networks, and computing centers. During scheduling, priority is given to using idle space-based resources (such as the Ka-band of relay satellites); 15% of the ground-based fiber optic bandwidth is reserved for cross-domain tasks to prevent link congestion. For example: A remote sensing satellite needs to urgently transmit data back, and it coordinates with the ground-based Beijing station through a space-based relay satellite (Ka-band) to complete the data transmission within 5 minutes.

[0131] The present invention also adopts a two-level conflict resolution mechanism to dynamically resolve conflicts and optimize resource utilization. Among them, the first-level scheduling generates candidate resources based on the task visibility window and domain load, and dynamically calculates the conflict degree through the bitmap decision matrix to shorten the conflict detection response time; the second-level scheduling adopts a local conflict resolution matrix, combined with the task matching value and conflict depth weighted calculation to reduce the resource conflict rate; moreover, the space-based Ku-band resources are forcibly reserved for fast-response tasks, and the optical fiber bandwidth of cross-domain collaborative tasks is pre-allocated to dynamically release idle resources and improve resource utilization.

[0132] In some embodiments, the method for dynamic scheduling and management of space measurement and control mission resources further includes establishing a dynamic resource reservation pool mechanism, wherein the dynamic resource reservation pool mechanism includes:

[0133] Mandatory reservation of no less than two space-based Ku-band beam resources for fast-response missions;

[0134] Pre-allocate 15% of the fiber link bandwidth between ground and base stations for cross-domain collaborative tasks;

[0135] Reserve resources by Exponential decay release, where R release is the amount of reserved resources that should be released at the current moment, R reserved is the total amount of initially reserved resources, β is the dynamic recovery coefficient, and the β value is dynamically adjusted in the range of 0.1-0.5 according to the urgency of the task. β = 0.5 for fast-response tasks and β = 0.2 for regular tasks. e is the base of the natural logarithm, t is the current time, and t0 is the start time of resource reservation.

[0136] In the cross-domain two-level scheduling strategy of the present invention, the first-level scheduling generates candidate resources based on the resource topology map, and supports the four-layer and eight-category dynamic division of space-based / ground-based resources; the second-level scheduling realizes multi-domain resource collaboration through the conflict resolution matrix, improves the success rate of cross-domain task scheduling, adopts the short message drive mode, shortens the emergency task scheduling delay, adjusts the weight through the dynamic attenuation factor when the burst task is inserted, and improves the success rate of high-priority task preemption.

[0137] It should be noted that the above figures are merely illustrative of the processes included in the method according to exemplary embodiments of the present invention and are not intended to be limiting. It is readily understood that the processes illustrated in the above figures do not indicate or limit the temporal order of these processes. Furthermore, it is readily understood that these processes may be executed synchronously or asynchronously, for example, in multiple modules.

[0138] It should be understood that, although the above is described in a certain order, these steps are not necessarily performed in sequence according to the above order. Unless clearly stated herein, the execution of these steps does not have strict order restrictions, and these steps can be performed in other orders. Moreover, a part of the steps of the present embodiment may include multiple steps or multiple stages, and these steps or stages are not necessarily performed at the same time, but can be performed at different times, and the execution order of these steps or stages is not necessarily performed in sequence, but can be performed in turn or alternately with at least a portion of the steps or stages in other steps or other steps.

[0139] See also Figure 3 As shown, an embodiment of the present invention further provides a space tracking and control mission resource dynamic scheduling and management system, the space tracking and control mission resource dynamic scheduling and management system includes:

[0140] Satellite-ground IoT terminal 100, used to collect satellite attitude, power supply voltage and radio frequency power data in real time, with a sampling frequency of ≥10Hz;

[0141] The digital twin construction module 200 is connected to the satellite-ground IoT terminal 100 to build a dynamic mapping system including a satellite orbit prediction model and an equipment health assessment model;

[0142] An intelligent scheduling engine 300, comprising a distributed DQN network and a particle swarm algorithm hybrid computing unit, configured to execute the aforementioned method for dynamic scheduling and management of space tracking and control mission resources;

[0143] The cross-domain scheduling coordinator 400 is used to implement a two-level scheduling strategy, including a first-level scheduler based on a resource topology map and a second-level scheduler based on a conflict resolution matrix;

[0144] Performance evaluation module 500, real-time calculation of task on-time completion rate η time , beam reuse index μ beam and maneuver cost coefficient C move and feeds back the evaluation results to the intelligent scheduling engine 300.

[0145] In this embodiment, the intelligent scheduling engine 300 includes:

[0146] The conflict resolution unit 301 is configured to mark the resource occupancy status in the bitmap and trigger the dynamic priority migration when it is detected that the time window overlap is ≥3 minutes;

[0147] The emergency response unit 302 is configured to complete resource preemption and rescheduling of existing tasks within 5 seconds upon receiving an emergency task with a priority level ≥ 2;

[0148] The fault-tolerant rescheduling unit 303 is configured to run two scheduling schemes, conservative and aggressive, in parallel, and automatically switch to the conservative scheme when the device health deviation is greater than 20%.

[0149] The digital twin construction module 200 includes:

[0150] The anomaly detection submodule 201 is configured to generate a level 3 alarm when the satellite battery temperature mutation rate is greater than 5°C / min;

[0151] Resource visualization submodule 202 dynamically displays the orbital distribution map of space-based resources and the heat map of ground-based resource load, with a refresh rate of ≥1 time / second;

[0152] The predictive maintenance submodule 203 predicts the remaining life of the equipment based on the LSTM network and starts the spare parts scheduling process when the predicted life is less than 30 days.

[0153] In this embodiment, the performance evaluation module 500 includes:

[0154] The three-dimensional measurement unit 501 calculates the task on-time completion rate:

[0155]

[0156] Among them, W ontime is the number of tasks completed on time, with an error of ≤±1 minute; W total is the total number of tasks;

[0157] Calculate the beam reuse index:

[0158]

[0159] in, is the actual usage time of a single beam; N beam is the total number of beams in the system; T total is the total duration of the statistical period;

[0160] Calculate the mobility cost coefficient:

[0161]

[0162] in, is the antenna azimuth adjustment value of the i-th task; is the antenna pitch angle adjustment value of the i-th task; P az Adjust the cost per azimuth unit; P el Adjusted cost for pitch angle units;

[0163] Mode switching unit 502, when n time <90% and μ beam When the task rate is >1.2, the batch scheduling ratio of routine tasks is increased to more than 75%;

[0164] Cost optimization unit 503, according to C move The value automatically selects the phased array radar combination solution with the lowest maneuvering cost.

[0165] In some embodiments, the aerospace tracking and control mission resource dynamic scheduling and management system further includes a dynamic resource reservation pool 600, and the dynamic resource reservation pool 600 includes:

[0166] Mandatory reservation of no less than two space-based Ku-band beam resources for fast-response missions;

[0167] Pre-allocate 15% of the fiber link bandwidth between ground and base stations for cross-domain collaborative tasks;

[0168] Reserve resources by Exponential decay release, where R release is the amount of reserved resources that should be released at the current moment, R reservedis the total amount of initially reserved resources, β is the dynamic recovery coefficient, and the β value is dynamically adjusted in the range of 0.1-0.5 according to the urgency of the task. β = 0.5 for fast-response tasks and β = 0.2 for regular tasks. e is the base of the natural logarithm, t is the current time, and t0 is the start time of resource reservation.

[0169] Through the above detailed steps, the aerospace measurement and control mission resource dynamic scheduling and management system of the present invention is used to execute the steps of the aerospace measurement and control mission resource dynamic scheduling and management method in the above embodiment, which will not be repeated here. The aerospace measurement and control mission resource dynamic scheduling and management method and system provided by the present invention systematically solve the core problems such as high-concurrency task conflicts, resource allocation delays and cross-domain collaboration difficulties in the field of aerospace measurement and control through task feature modeling, multi-algorithm fusion scheduling, dynamic resource management and cross-domain collaboration mechanism, significantly improving the real-time performance, reliability and resource utilization efficiency of the measurement and control system, providing key technical support for future large-scale constellation management and emergency measurement and control, supporting dynamic matching of resources between large-scale low-orbit satellites (such as Starlink and remote sensing constellations) and ground stations, improving measurement and control coverage, and also applicable to scenarios such as satellite failure rescue and space threat avoidance.

[0170] The above are exemplary embodiments disclosed in the present invention, but it should be noted that various changes and modifications may be made without departing from the scope of the embodiments disclosed in the claims. The functions, steps and / or actions of the method claims according to the disclosed embodiments described herein do not need to be performed in any particular order. In addition, although the elements disclosed in the embodiments of the present invention may be described or required in individual form, they may also be understood as multiple unless expressly limited to the singular.

[0171] It should be understood that, as used herein, the singular form "a" or "an" is intended to include the plural form as well, unless the context clearly supports an exception. It should also be understood that, as used herein, "and / or" refers to any and all possible combinations of one or more of the items listed in association. The serial numbers of the embodiments disclosed in the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0172] Those skilled in the art should understand that the discussion of any of the above embodiments is merely illustrative and is not intended to imply that the scope of the disclosure of the embodiments of the present invention (including the claims) is limited to these examples. Within the spirit of the embodiments of the present invention, the technical features of the above embodiments or different embodiments may be combined, and there are many other variations of different aspects of the above embodiments of the present invention, which are not provided in detail for the sake of simplicity. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the embodiments of the present invention should be included in the scope of protection of the embodiments of the present invention.

Claims

1. A method for dynamic scheduling and management of space measurement and control mission resources, characterized in that: The method comprises the following steps: Space tracking and control tasks are divided into routine tasks, fast-response tasks, and five types of complex-constraint tasks. Task feature vectors are defined based on the time sensitivity, resource demand intensity, and inter-task coupling of the tasks. Satellite health parameters, equipment operating conditions, and environmental data are collected in real time through satellite-ground IoT terminals to build a digital twin, and the real-time data is fed back to the intelligent scheduling engine. The intelligent scheduling engine combines a distributed DQN network and a particle swarm algorithm to prioritize tasks, match resources, and resolve conflicts based on task feature vectors, enabling global optimization and scheduling of space tracking and control tasks. Based on the bitmap decision matrix calculation, the task matching value and conflict depth are combined for weighted calculation to dynamically adjust the priority of conflicting tasks and dynamically resolve conflicts generated during the scheduling process. During global optimization scheduling, cross-domain two-level scheduling is performed. The first-level scheduling generates candidate resources based on the task visibility window and domain load, and the second-level scheduling generates the final scheduling plan through local conflict resolution, supporting the coordinated scheduling of space-based and ground-based resources.

2. The method for dynamic scheduling and management of aerospace measurement and control mission resources according to claim 1, wherein: The defined task feature vector is T = [τ, μ, σ], where τ is the time sensitivity of the task, which is the difference between the task deadline and the available time window; μ is the resource demand intensity of the task, which indicates the number of resources required for task execution; and σ is the coupling degree of the task, which indicates the interdependence between tasks.

3. The method for dynamic scheduling and management of aerospace measurement and control mission resources according to claim 2, wherein: In the task feature vector: Time sensitivity τ is the difference between the duration of the available time window and the remaining execution time of the task; The resource requirement intensity μ is the normalized value of the sum of the number of antenna beams and computing units required for task execution; The coupling degree σ is obtained by calculating the data dependency matrix between tasks. When there is equipment sharing or time window overlap, the σ value increases by 20%-50%.

4. The method for dynamic scheduling and management of aerospace measurement and control mission resources according to claim 1, wherein: The processing of the five types of compound constraint tasks includes: Synchronous tracking tasks: Using sliding time window scanning, the system searches for the intersection of available time windows of multiple devices with a preset step size. When the duration of the intersection is greater than or equal to the minimum task requirement, the system automatically locks the optimal start time. Multi-station relay missions: Implement a split strategy of prioritizing difficult tasks before easy ones, prioritize the allocation of scarce Ka-band resources, and allow mission arcs to dynamically adjust time windows within a ±15-second range. Generate a network replenishment plan to support mission arc preemption and replacement. Cross-domain collaborative tasks: Through a two-level scheduling process, the first-level scheduling generates candidate resources based on the visible window and domain load, and the second-level scheduling generates the final solution through local conflict resolution. In the first-level scheduling stage, at least 15% of the fiber link bandwidth resources are reserved; Emergency rescue tasks: A dynamic preemption strategy is adopted to allow tasks with a priority level ≥ 2 to interrupt low-coupling regular tasks; Multi-objective joint task: Through the resource time-sharing multiplexing mechanism, the task is split into multiple sub-stages and allocated independent resource pools.

5. The method for dynamic scheduling and management of space measurement and control mission resources according to claim 4, characterized in that: The distributed DQN network adopts a dual-channel feature fusion mechanism, including: The spatial channel uses a graph convolutional network to model the topological relationship between space-based and ground-based resources. The nodes include satellite orbit altitude, ground station longitude and latitude, and link attenuation parameters. The time series channel uses LSTM to capture the dynamic evolution characteristics of the task queue and updates the hidden state every 60 seconds; The dual-channel weights are dynamically balanced through the attention mechanism, and the spatial channel dominant mode is activated when the resource conflict rate is >30%.

6. The method for dynamic scheduling and management of space tracking and control mission resources according to claim 3, characterized in that: Based on the bitmap decision matrix calculation, when combining the task matching value and conflict depth for weighted calculation, the task matching value M j and conflict depth C i The weighted formula is: Among them, D ij Indicates the scheduling conflict between task i and task j. The larger the value, the higher the conflict degree, which needs to be resolved first. i Indicates the conflict depth of task i, which indicates the degree of conflict between task i and other tasks during the scheduling process. The greater the conflict depth, the more priority the task needs to be scheduled; M j represents the matching value of task j, which indicates the matching degree between task j and resources; T i represents the timeliness of task i; ω1, ω2, ω3 represent the weighted coefficients of conflict depth, task matching degree and timeliness respectively; λ is a dynamic attenuation factor, which represents the impact of task timeliness on scheduling priority. When task timeliness T i When the threshold is reached, λ=1, otherwise, λ=0.

5.

7. The method for dynamic scheduling and management of space tracking and control mission resources according to claim 6, characterized in that: When prioritizing tasks, matching resources, and resolving conflicts based on task feature vectors, a short message-driven mode is adopted to complete resource matching and conflict resolution within 10 seconds. The emergency task scheduling delay is ≤5 seconds. When sudden tasks are inserted, the resource allocation weight is adjusted through the dynamic attenuation factor λ to prioritize the execution of high-priority tasks.

8. The method for dynamic scheduling and management of space measurement and control mission resources according to claim 1, wherein: The coordinated scheduling of space-based and ground-based resources includes: Categorize tracking and control resources into two categories: space-based and ground-based, and four status categories: planned, idle, transferable, and preemptible, for a total of eight resource combinations. Priority will be given to applying for idle space-based resources and ground-based resources with minimal conflicts. Priority migration and cross-domain conflict resolution are performed on conflicting tasks, supporting concurrent scheduling of multiple users and multiple tasks.

9. The method for dynamic scheduling and management of space tracking and control mission resources according to claim 7, wherein: The method for dynamic scheduling and management of space measurement and control mission resources further includes establishing a dynamic resource reservation pool mechanism, wherein the dynamic resource reservation pool mechanism includes: Mandatory reservation of no less than two space-based Ku-band beam resources for fast-response missions; Pre-allocate 15% of the fiber link bandwidth between ground and base stations for cross-domain collaborative tasks; Reserve resources by Exponential decay release, where R release is the amount of reserved resources that should be released at the current moment, R reserved is the total amount of initially reserved resources, β is the dynamic recovery coefficient, and the β value is dynamically adjusted in the range of 0.1-0.5 according to the urgency of the task. β = 0.5 for fast-response tasks and β = 0.2 for regular tasks. e is the base of the natural logarithm, t is the current time, and t0 is the start time of resource reservation.

10. A dynamic scheduling and management system for aerospace measurement and control mission resources, characterized in that: Includes the following components: Satellite-ground IoT terminal, used to collect satellite attitude, power supply voltage and RF power data in real time, with a sampling frequency of ≥10Hz; A digital twin construction module connects the satellite-ground IoT terminals to build a dynamic mapping system including a satellite orbit prediction model and an equipment health assessment model; An intelligent scheduling engine, comprising a distributed DQN network and a particle swarm algorithm hybrid computing unit, configured to execute the method for dynamic scheduling and management of aerospace measurement and control mission resources according to any one of claims 1 to 9; A cross-domain scheduling coordinator is used to implement a two-level scheduling strategy, including a first-level scheduler based on a resource topology map and a second-level scheduler based on a conflict resolution matrix; Performance evaluation module, real-time calculation of task on-time completion rate η time , beam reuse index μ beam and maneuver cost coefficient C move and feeds the evaluation results back to the intelligent scheduling engine.

Images