Multi-frequency task-oriented satellite-ground integrated task planning method

Abstract

The invention discloses a satellite-ground integrated task planning method for multi-frequency tasks, and belongs to the field of satellite task planning. The method specifically comprises the following steps: firstly, receiving an aerospace observation grid element task set to form a grid element task pool; secondly, a ground task planning system comprehensively considers satellite electric quantity, storage and the like to preliminarily screen grid element tasks, and a single-satellite grid element task set is formed; secondly, uploading the single-satellite grid element tasks to each satellite platform through a measurement and control link; and finally, the satellite platform performs access calculation on the satellite based on an actual operation orbit to obtain accurate grid access time, namely load startup and shutdown time, and then single-satellite task planning is completed based on an improved genetic algorithm to form a final observation scheme. According to the invention, joint task planning is carried out by adopting satellite-ground resources, and a sectional task planning mode of obtaining a redundant observation scheme through ground rough planning and obtaining a final observation scheme through satellite fine planning is innovatively proposed, so that the observation coverage rate and task guarantee timeliness are effectively improved.

Description

Technical Field

[0001] This invention belongs to the field of satellite mission planning, specifically relating to a space-ground integrated mission planning method for multi-frequency missions, used to realize segmented space-ground mission planning. Background Technology

[0002] For a long time, satellite mission planning has primarily relied on ground systems. This is a centralized architecture: the ground control center receives user requests, comprehensively considers factors such as satellite orbit, payload constraints, and weather, determines a detailed satellite mission plan through a large-scale computer mechanism, and then uploads the command sequence to the satellite using scarce ground resources, which then passively executes the commands. This mission planning method can no longer meet the current situation of Earth observation with the surge in the number of satellites in orbit and the increasingly complex application requirements. Furthermore, it is difficult to perceive the precise status of satellites in real time (such as battery power and storage), and the planning tends to be conservative, resulting in the underutilization of the satellite's valuable overflight window and payload resources.

[0003] With the development of satellite technology and the continuous enhancement of onboard payload computing power, domestic and international research institutions have gradually begun to pay attention to the research of onboard autonomous mission planning. However, current onboard autonomous mission planning knowledge serves as an auxiliary to ground-based centralized mission planning, registering a small number of target points on the satellite. Onboard autonomy is only triggered when satellite and ground telemetry resources are not visible for extended periods or when the satellite has been idle for too long. A massive number of targets are still planned using traditional centralized mission planning, resulting in low utilization of onboard computing resources and inaccurate planning schemes. Therefore, there is an urgent need for a space-ground integrated joint mission planning method that fully utilizes onboard resources and generates accurate satellite mission planning schemes based on the real-time status of satellite power and storage, ensuring the efficient completion of user requirements. Summary of the Invention

[0004] To address the problems of ground-based centralized mission planning failing to perceive the onboard power and storage status in real time, resulting in conservative planning and an inability to maximize the utilization of onboard computing and observation resources, this invention proposes a space-ground integrated mission planning method for multi-frequency missions.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0006] A space-ground integrated mission planning method for multi-frequency missions includes the following steps:

[0007] Step 1: Receive the set of space observation grid meta-tasks to form a grid meta-task pool. Each grid meta-task includes grid code, observation frequency, grid priority, list of observed satellite codes, observation start time and observation end time.

[0008] Step 2: The ground mission planning system performs preliminary screening of grid meta-tasks and, in conjunction with grid meta-task information and satellite status information, allocates grid meta-tasks to each satellite to form a single-satellite grid meta-task set.

[0009] Step 3: Upload the single-satellite grid meta-tasks to each satellite platform via the telemetry, tracking, and command (TT&C) link;

[0010] Step 4: After receiving the single-satellite grid meta-task set, the satellite platform performs access calculations on-board based on the actual operating orbit to obtain the precise grid access time, i.e., payload power-on / off time. Then, based on the improved genetic algorithm, it completes the single-satellite mission planning to form the final observation scheme.

[0011] Furthermore, the specific steps of step 2 are as follows:

[0012] Step 2.1: Divide the grid meta-tasks according to the grid code, and sort the grids in descending order according to grid priority;

[0013] Step 2.2: Starting with the highest priority grid, proceed according to the grid's observation frequency. Grid demand observation cycle Divide into equal intervals, each time interval being... Among them, the observation frequency of the grid This represents the highest observation frequency of the grid meta-task within the grid.

[0014] Step 2.3: Calculate the remaining percentage of satellite battery power and storage, and normalize the battery power and storage percentages accordingly. ; In the formula, This represents the normalized results for power consumption and storage. This indicates the satellite's current remaining battery power. This indicates the satellite's full charge level. Indicates the satellite's current remaining storage space. Indicates the satellite's maximum storage space. and These are represented by the weights of power and storage allocation, respectively. ; Then to Sort in descending order and optimize within each time interval. The largest satellite is assigned a grid cell task, and for each grid cell task assigned, the corresponding satellite's power and storage are reduced. and Re-updated Sort in descending order; where, and These represent the reduction values ​​for power consumption and storage, respectively, and can be set according to requirements;

[0015] Step 2.4, further optimize the distance from the start time within each time interval. The most recent satellite assignment corresponds to the corresponding grid meta-task; after all time intervals are assigned, determine the grid requirement observation cycle. Number of grid meta-tasks already assigned Does r≥ If so, proceed to step 2.5; otherwise, randomly select a grid meta-task within the grid and assign it to the satellite, such that r ≥ Proceed to step 2.5;

[0016] Step 2.5, return to step 2.2 until all grids are processed, finally forming a single-star grid meta-task set.

[0017] Furthermore, the specific steps of step 4 are as follows:

[0018] Step 4.1: After receiving the single-satellite grid meta-task set, the satellite platform re-performs the access calculation on-board based on the actual operating orbit of the satellite to obtain the accurate payload power-on and power-off times.

[0019] Step 4.2 involves performing structured preprocessing on the grid meta-task information, including grid priority, satellite designation, and payload power-on / off times, to generate a gene pool. The size of the gene pool is related to the number of grids. A complete match is achieved, with each gene location storing the core information of its corresponding grid, including grid encoding, spatial location, and a list of grid metatasks. The maximum length of genes in the gene pool is assumed to be... That is, the grid corresponding to this gene contains For each grid meta-task, the gene pool is represented as:

[0020]

[0021] in, Indicates the first The first gene locus at the _th There are 1 grid metatasks, compiled using integers starting from 1. For grid metatasks with a number less than 1, The empty positions of the gene loci are filled with 0s in the matrix;

[0022] Step 4.3: Generate chromosomes from the gene pool. The length of each chromosome is the same as the length of the gene pool. Each gene position on the chromosome represents a grid meta-task. The length of the gene pool is the number of grids.

[0023] Step 4.4: Input the structured grid-based task sequence into the improved genetic algorithm for iterative task optimization. During task planning, grid coverage is considered. and As the direction of optimization; the objective function is: ; Grid coverage optimization: Grid coverage The evaluation is quantified through the "sum of priorities," specifically by statistically analyzing the grids covered by the planning task set, summing the preset priorities of each grid, and forming a comprehensive coverage score; mathematically expressed as: ; in, This represents grid coverage, i.e., the sum of priorities. The set of grids covered by the planning task. For grid The preset priority is used; a deduplication mechanism is employed to ensure that the priority of the same grid is not calculated repeatedly, and the optimization objective is to maximize the priority. ; Task timeliness optimization: Task timeliness Focusing on the matching accuracy between task execution time and demand window, specifically: for each observation grid, first determine the start time of the earliest task scheduled for it in the plan, then associate it with the corresponding demand time window of that grid, and quantify it by calculating the average of the time deviation percentages; where the demand time window includes the demand start time and end time; mathematically expressed as: ; in, This represents the overall quantitative result of task timeliness. Total number of grid cells; For the first The earliest task start time for each grid and The first The start and end times of the requirements for each grid; the optimization objective is to minimize them. ; Finally, the selection is based on the objective function. The chromosome with the highest value.

[0024] Compared with the prior art, the present invention has the following advantages:

[0025] (1) The satellite mission planning constructed in this invention adopts a segmented mission planning approach. After receiving the observation mission, the ground performs a preliminary coarse planning, giving full play to the advantage of sufficient ground computing power to complete the access calculation and task allocation for massive observation needs, and obtain a redundant single-satellite grid meta-task set. The satellite performs fine planning, senses the satellite's storage and power status in real time, and resolves conflicts based on the satellite's real-time orbital information, and finally obtains a mission planning scheme that meets the requirements, effectively shortening the mission planning time while improving the accuracy of mission planning;

[0026] (2) This invention considers the balanced and efficient execution of multiple tasks. It divides the required time period into equal intervals according to the observation frequency. At the same time, it prioritizes the selection of grid meta-tasks that are closer to the start time of the time interval on the ground. On the satellite, it uses timeliness as one of the optimization directions of the improved genetic algorithm. This avoids the excessive concentration of multiple tasks while effectively ensuring the timeliness of the tasks. Attached Figure Description

[0027] Figure 1 This is a system flowchart of the present invention.

[0028] Figure 2 This is a schematic diagram of the ground mission planning of the present invention.

[0029] Figure 3 The flowchart of the improved genetic algorithm of this invention is shown.

[0030] Figure 4 This is a schematic diagram of the gene bank of the present invention.

[0031] Figure 5 This is a schematic diagram of the chromosome in this invention.

[0032] Figure 6 This is a schematic diagram illustrating the optimized selection of the solution in this invention. Detailed Implementation

[0033] The invention's concept, technical advantages, and resulting technical effects will be clearly and completely described below with reference to the accompanying drawings and specific embodiments, so as to fully understand the invention's purpose, features, and effects. It should be noted that the specific embodiments described herein are only for explaining the invention and do not limit the invention.

[0034] This invention proposes a space-ground integrated mission planning method for multi-frequency missions. This method leverages the advantages of abundant ground computing resources and the ability of satellites to perceive resource status in real time, and completes the planning of massive observation missions in a segmented manner to obtain accurate satellite mission planning schemes.

[0035] Reference Figure 1 This invention first acquires a space observation grid meta-task set, performs preliminary screening on the ground to form a single-satellite grid meta-task set, and uploads it to each satellite platform via the telemetry, tracking, and command (TT&C) link. Onboard, an improved genetic algorithm is used for refined single-satellite mission planning to form the final observation scheme. The specific steps of this method include:

[0036] Step 1: Receive the set of space observation grid meta-tasks to form a grid meta-task pool. Each grid meta-task contains information such as grid code, observation frequency, grid priority, list of observed satellite codes, observation start time, and observation end time.

[0037] Step 2: The ground mission planning system performs preliminary screening of grid meta-tasks and, in conjunction with mission information such as priority and observation frequency of grid meta-tasks, as well as status information such as satellite power and storage, allocates grid meta-tasks to each satellite to form a single-satellite grid meta-task set.

[0038] Specifically, the steps for ground mission planning are as follows: Figure 2 As shown:

[0039] Step 2.1: Divide the grid meta-tasks according to the grid code, and sort the grids in descending order according to grid priority;

[0040] Step 2.2: Starting with the highest priority grid, proceed according to the grid's observation frequency. Grid demand observation cycle Divide into equal intervals, each time interval being... Among them, the observation frequency of the grid This represents the highest observation frequency of the grid meta-task within the grid.

[0041] Step 2.3: Calculate the remaining percentage of satellite battery power and storage, and normalize the battery power and storage percentages accordingly. ; In the formula, This represents the normalized results for power consumption and storage. This indicates the satellite's current remaining battery power. This indicates the satellite's full charge level. Indicates the satellite's current remaining storage space. Indicates the satellite's maximum storage space. and These are represented by the weights of power and storage allocation, respectively. ; Then to Sort in descending order and optimize within each time interval. The largest satellite is assigned a grid cell task, and for each grid cell task assigned, the corresponding satellite's power and storage are reduced. and Re-updated Sort in descending order; where, and These represent the reduction values ​​for power consumption and storage, respectively, and can be set according to requirements;

[0042] Step 2.4, further optimize the distance from the start time within each time interval. The most recent satellite assignment corresponds to the corresponding grid meta-task; after all time intervals are assigned, determine the grid requirement observation cycle. Number of grid meta-tasks already assigned Does r≥ If so, proceed to step 2.5; otherwise, randomly select a grid meta-task within the grid and assign it to the satellite, such that r ≥ Proceed to step 2.5;

[0043] Step 2.5: Return to step 2.2 until all grids are processed, ultimately forming a single-star grid meta-task set;

[0044] It should be noted that the number of grid meta-tasks is greater than the observation frequency in order to leave room for conflict resolution in the planning of subsequent on-board missions and to increase the number of redundant observations in order to achieve better observation results.

[0045] Step 3: Upload the single-satellite grid meta-task to each satellite platform via the telemetry, tracking, and command (TT&C) link.

[0046] Step 4: After receiving the single-satellite grid meta-task set, the satellite platform performs access calculations on-board based on the actual operating orbit to obtain the precise grid access time, i.e., payload power-on / off time. Then, based on the improved genetic algorithm, it completes the single-satellite mission planning to form the final observation scheme.

[0047] Specifically, the single-satellite mission planning steps based on the improved genetic algorithm are as follows: Figure 3 As shown:

[0048] (4.1) After receiving the single-satellite grid meta-task set, the satellite platform re-performs the access calculation on the satellite based on the actual operating orbit of the satellite to obtain the accurate payload power-on and power-off time;

[0049] (4.2) Structured preprocessing of grid meta-task information such as grid priority, satellite designation, and payload start-up / shutdown time is performed to generate a gene pool. The size of the gene pool is related to the number of grids. A perfect match is achieved, with each gene locus precisely storing the core information of its corresponding grid, including grid encoding, spatial location, and a list of included meta-tasks. It is assumed that the maximum length of genes in the gene pool is... That is, the grid corresponding to this gene contains For each meta-task, the gene pool can be represented as:

[0050]

[0051] in Indicates the first The first gene locus at the _th Each metatask is compiled using integers starting from 1 (metatasks within the same grid are numbered sequentially, and numbers from different grids do not interfere with each other). For metatasks with fewer than [number missing], [the following is unclear due to incomplete text]. The empty positions of the gene loci are filled with 0 in the matrix.

[0052] For example, if a satellite needs to observe three grids, A, B, and C, where grid A contains 3 meta-tasks, grid B contains 2 meta-tasks, and grid C contains 4 meta-tasks, then the gene pool size is 3 (consistent with the number of grids), and the meta-task numbers for grid A are 1, 2, and 3, for grid B they are 1 and 2, and for grid C they are 1, 2, 3, and 4. Figure 4 As shown;

[0053] (4.3) Generate chromosomes based on the gene pool from the previous step. The length of each chromosome is the same as the length of the gene pool (i.e., the number of grids). Each gene position on the chromosome represents a grid meta-task.

[0054] For example, if the coding sequence of a chromosome is "1, 0, 3", it means that grid A selects the first meta-task in its meta-task list, grid B selects no meta-task, and grid C selects the third meta-task in its meta-task list. This encoding method directly maps the meta-task selection results of each grid, providing a clear decision-making basis for subsequent task conflict resolution, such as... Figure 5 As shown;

[0055] (4.4) Input the structured grid-based task sequence into the improved genetic algorithm for iterative optimization of the task, such as... Figure 6 As shown in the diagram, each number forms a chromosome. During task planning, grid coverage is considered. and As the direction of optimization; the objective function is: ; Grid coverage optimization: Grid coverage The evaluation is quantified through the "sum of priorities," specifically by statistically analyzing the grids covered by the planning task set, summing the preset priorities of each grid, and forming a comprehensive coverage score; mathematically expressed as: ; in, This represents grid coverage, i.e., the sum of priorities. The set of grids covered by the planning task. For grid The preset priority is used; a deduplication mechanism is employed to ensure that the priority of the same grid is not calculated repeatedly, and the optimization objective is to maximize the priority. ; Task timeliness optimization: Task timeliness Focusing on the matching accuracy between task execution time and demand window, specifically: for each observation grid, first determine the start time of the earliest task scheduled for it in the plan, then associate it with the corresponding demand time window of that grid, and quantify it by calculating the average of the time deviation percentages; where the demand time window includes the demand start time and end time; mathematically expressed as: ; in, This represents the overall quantitative result of task timeliness. Total number of grid cells; For the first The earliest task start time for each grid and The first The start and end times of the requirements for each grid; the optimization objective is to minimize them. ; Finally, the selection is based on the objective function. The chromosome with the highest value.

Claims

1. A space-ground integrated mission planning method for multi-frequency missions, characterized in that, Includes the following steps: Step 1: Receive the set of space observation grid meta-tasks to form a grid meta-task pool. Each grid meta-task includes grid code, observation frequency, grid priority, list of observed satellite codes, observation start time and observation end time. Step 2: The ground mission planning system performs preliminary screening of grid meta-tasks and, in conjunction with grid meta-task information and satellite status information, allocates grid meta-tasks to each satellite to form a single-satellite grid meta-task set. Step 3: Upload the single-satellite grid meta-tasks to each satellite platform via the telemetry, tracking, and command (TT&C) link; Step 4: After receiving the single-satellite grid meta-task set, the satellite platform performs access calculations on-board based on the actual operating orbit to obtain the precise grid access time, i.e., payload power-on / off time. Then, based on the improved genetic algorithm, it completes the single-satellite mission planning to form the final observation scheme.

2. The space-ground integrated mission planning method for multi-frequency missions according to claim 1, characterized in that, Step 2 is detailed below: Step 2.1: Divide the grid meta-tasks according to the grid code, and sort the grids in descending order according to grid priority; Step 2.2: Starting with the highest priority grid, proceed according to the grid's observation frequency. Grid demand observation cycle Divide into equal intervals, each time interval being... Among them, the observation frequency of the grid This represents the highest observation frequency of the grid meta-task within the grid. Step 2.3: Calculate the remaining percentage of satellite battery power and storage, and normalize the battery power and storage percentages accordingly. ; In the formula, This represents the normalized results for power consumption and storage. This indicates the satellite's current remaining battery power. This indicates the satellite's full charge level. Indicates the satellite's current remaining storage space. Indicates the satellite's maximum storage space. and These are represented by the weights of power and storage allocation, respectively. ; Then to Sort in descending order and optimize within each time interval. The largest satellite is assigned a grid cell task, and for each grid cell task assigned, the corresponding satellite's power and storage are reduced. and Re-updated Sort in descending order; where, and These represent the reduction values ​​for power consumption and storage, respectively, and can be set according to requirements; Step 2.4, further optimize the distance from the start time within each time interval. The most recent satellite assignment corresponds to the corresponding grid meta-task; after all time intervals are assigned, determine the grid requirement observation cycle. Number of grid meta-tasks already assigned Does r≥ If so, proceed to step 2.5; otherwise, randomly select a grid meta-task within the grid and assign it to the satellite, such that r ≥ Proceed to step 2.5; Step 2.5, return to step 2.2 until all grids are processed, finally forming a single-star grid meta-task set.

3. The space-ground integrated mission planning method for multi-frequency missions according to claim 1, characterized in that, Step 4 is detailed below: Step 4.1: After receiving the single-satellite grid meta-task set, the satellite platform re-performs the access calculation on-board based on the actual operating orbit of the satellite to obtain the accurate payload power-on and power-off times. Step 4.2 involves performing structured preprocessing on the grid meta-task information, including grid priority, satellite designation, and payload power-on / off times, to generate a gene pool. The size of the gene pool is related to the number of grids. A complete match is achieved, with each gene location storing the core information of its corresponding grid, including grid encoding, spatial location, and a list of grid metatasks. The maximum length of genes in the gene pool is assumed to be... That is, the grid corresponding to this gene contains For each grid meta-task, the gene pool is represented as: ； in, Indicates the first The first gene locus at the _th There are 1 grid metatasks, compiled using integers starting from 1. For grid metatasks with a number less than 1, The empty positions of the gene loci are filled with 0s in the matrix; Step 4.3: Generate chromosomes from the gene pool. The length of each chromosome is the same as the length of the gene pool. Each gene position on the chromosome represents a grid meta-task. The length of the gene pool is the number of grids. Step 4.4: Input the structured grid-based task sequence into the improved genetic algorithm for iterative task optimization. During task planning, grid coverage is considered. and As the direction of optimization; the objective function is: ; Grid coverage optimization: Grid coverage The evaluation is quantified through the "sum of priorities," specifically by statistically analyzing the grids covered by the planning task set, summing the preset priorities of each grid, and forming a comprehensive coverage score; mathematically expressed as: ； in, This represents grid coverage, i.e., the sum of priorities. The set of grids covered by the planning task. For grid The preset priority is used; a deduplication mechanism is employed to ensure that the priority of the same grid is not calculated repeatedly, and the optimization objective is to maximize the priority. ; Task timeliness optimization: Task timeliness Focusing on the matching accuracy between task execution time and demand window, specifically: for each observation grid, first determine the start time of the earliest task scheduled for it in the plan, then associate it with the corresponding demand time window of that grid, and quantify it by calculating the average of the time deviation percentages; where the demand time window includes the demand start time and end time; mathematically expressed as: ； in, This represents the overall quantitative result of task timeliness. Total number of grid cells; For the first The earliest task start time for each grid and The first The start and end times of the requirements for each grid; the optimization objective is to minimize them. ; Finally, the objective function is used to select... The chromosome with the highest value.

Images