A multi-satellite task planning method and system for large-area targets

Abstract

The application provides a multi-satellite task planning method and system for a large-area target, which strips the target area to obtain a strip set of satellite resources. Then, the target area is gridded to find a grid set covered by the strips, and the area of the target area covered by the grid is calculated. A planning model is established with the target of maximizing the strip coverage of the target area and minimizing the resource usage. Finally, a deep adaptive large neighborhood search algorithm is proposed to solve the problem. Due to the simple structure of the planning model, a planning scheme for observing the regional target with high coverage rate by using as few satellite resources as possible can be quickly obtained.

Description

Technical Field

[0001] This invention belongs to the field of mission planning and scheduling, and in particular relates to a multi-satellite mission planning method and system for large-area targets. Background Technology

[0002] Imaging satellites are among the most valuable spacecraft. With the development of science and technology, satellite imaging technology has gradually matured. However, the demands on satellite observation missions have also become more complex and diverse, primarily encompassing fields such as military warfare, disaster relief, and environmental monitoring. Consequently, mission types have evolved from early single-point target observation, small-area strip observation, and reciprocating single-point target observation to moving target observation, large-area target observation, and multi-area target observation.

[0003] Large-area targets refer to targets that a single satellite cannot effectively utilize its swath width for observation, requiring multiple satellites to collaborate in imaging to cover them. For example, in the MH370 incident, 21 satellites worked together to monitor and search for the affected area. Currently, multi-satellite collaboration has become a means to accomplish complex or large-scale missions, while also maximizing satellite efficiency. However, given the demands of large-area imaging missions and limited satellite resources (limited number of resources, battery capacity, and storage capacity), how to efficiently cover regional targets with as few satellite resources as possible remains a problem that needs to be solved.

[0004] In practice, such as Figure 1 Because different missions have different characteristics, the shape and size of the areas to be observed vary. Therefore, how to combine satellite strips to maximize system benefits is crucial. Furthermore, the different deflection angles and on / off times of each satellite also generate multiple strips, leading to various combinations of observation strips. This results in enormous computational space complexity, making the problem NP-hard. Therefore, for multi-satellite mission planning problems involving large-area targets, a large-area target coverage method based on multi-strip stitching is a vital technique in the entire planning process. In addition, traditional methods mostly pre-divide the regional target into several strips based on satellite trajectories before planning, which is a segmented mission planning method. This is not conducive to global optimization in large-area target planning scenarios with wide areas and many visible satellite strips, necessitating high-performance models and methods to meet the observation needs of large-area targets. Summary of the Invention

[0005] The technical problem to be solved by this invention is how to use as few satellite resources as possible to efficiently cover regional targets. It proposes a multi-satellite mission planning method and system for large-area targets.

[0006] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows:

[0007] A multi-satellite mission planning method for large-area targets includes the following steps:

[0008] Step 1: Obtain the target set of the region And satellite resource set S;

[0009] Step 2: Calculate the target set for the region Each target task strip set i To obtain the regional target set The set of stripes B for all target tasks in the middle, Represents the target set of the region The i-th task, , The number of target tasks in the regional target set;

[0010] Step 3: Calculate the regional target set based on the strip set B of all target tasks. Coverage of each target area;

[0011] Step 4: Based on the coverage of regional targets, construct a multi-satellite scheduling task planning model for large areas;

[0012] Step 5: Solve the planning model for large-area multi-satellite scheduling tasks and output the planning scheme.

[0013] Furthermore, in step 3, the regional target set is calculated. The method for determining the coverage of each target area is as follows:

[0014] Step 3.1: Perform gridding on the regions where all targets are located in the target set R to obtain the coordinates of each grid in the target region after gridding;

[0015] Step 3.2: Based on the coordinates of the grid, determine whether all grids containing the target area are covered by the stripe, and count the number of all grids covered by the satellite stripe;

[0016] Step 3.3: Calculate the percentage of the target area in each grid to obtain the target area coverage.

[0017] Furthermore, the planning model for large-area multi-satellite scheduling tasks in step 4 is:

[0018] The objective function is:

[0019]

[0020] In equation (1), The primary objective function is to maximize task coverage. It is a secondary objective function, which minimizes the number of stripes used; it only applies if the function... The optimization objective reaches the threshold After that, the function Only then will it be related to the function Simultaneously optimized;

[0021] In equation (2), The parameter is 0-1, indicating whether the k-th grid is covered by the stripe (1 for yes, 0 for no); m represents the number of grid cells. Indicates the first The area of ​​the target region existing in each grid. This represents the total area of ​​the target region.

[0022] In equation (3), Let 'a' be a 0-1 decision variable, representing the first... The first satellite The first transit Whether a strip is used, 1 for yes, 0 for no; This represents the total number of transits within the planning period, where n is the number of satellites participating in the planning. Indicates the first satellite In the A collection of strips of goods that cross the border. Indicates the first satellite In the Total number of stripes per transit

[0023] The constraints are:

[0024]

[0025] Equation (4) represents the attitude transition relationship constraint between strips during a single transit of the same satellite; Indicates the first The first satellite The first in the strip set of the next transit The end time of each strip; Indicates stripe and strips Conversion time between Indicates stripe and strips The total angle of the conversion between them;

[0026]

[0027] Indicates the first The first satellite The first in the strip set of the next transit Each stripe Indicates the first The first satellite The first in the strip set of the next transit Each stripe Indicates the first The first satellite The first in the strip set of the next transit The start time of each strip; The parameter is 0-1, indicating the first... The first satellite The first in the strip set of the next transit The first stripe and the first All stripes are used, 1 for yes, 0 for no; Represents the i-th satellite's... A collection of strips that cross the border at a time;

[0028] Equation (5) indicates that both stripes must be selected simultaneously for the strip transition time to be considered;

[0029] Equation (6) represents the upper limit constraint on the energy used by each satellite. : indicates satellite The maximum energy available per lap; Indicates the first The first satellite The first in the strip set of the next transit The energy required for each strip.

[0030] Furthermore, in step 2, the target set of the region is calculated. Each target task strip set i To obtain the regional target set The method for determining the strip set B of all target tasks is:

[0031] Step 2.1: Based on each task Based on the location information and observation time information, satellites are selected sequentially. , h =1,2…|S|, where |S| represents the number of satellites in the satellite resource set S. The satellites are calculated... For the task Observation strip set Finally, the task was obtained. All satellite strip sets ,

[0032] ;

[0033] Step 2.2: Merge all satellite strip sets for all missions to obtain a strip set for large-area targets. .

[0034] Furthermore, the method for solving the planning model of multi-satellite scheduling tasks for large areas is a deep adaptive large neighborhood search algorithm, specifically:

[0035] Step 5.1: Randomly generate an initial population, assign each individual in the initial population to a predefined thread as an initial solution, and input the algorithm library and operator library, as well as the algorithm and operator parameters, into each thread;

[0036] Step 5.2: Set the algorithm running time. Each thread selects an algorithm from the algorithm library by roulette wheel. At the same time, during the running of each algorithm, operators are selected from the operator library by roulette wheel. Each thread performs a search based on the selected algorithm and operator and records the historical optimal solution of the algorithm. The historical optimal solution refers to the solution with the highest objective function value among the current solution and historical solutions.

[0037] Step 5.3: Merge the historical best solutions of each algorithm to form the current best solution set; retain the solutions with the highest objective function values ​​in the current best solution set. Each individual constructs the current population and calculates the contribution of each algorithm and operator to the current population, and updates the probability of each algorithm and operator being selected based on the contribution.

[0038] Step 5.4: Randomly pair individuals in the current population into two pairs, run the crossover operator to obtain a new population, and reallocate the initial solution to each thread in a roulette wheel manner according to the objective function value;

[0039] Step 5.5: If the termination condition is met, output the optimal individual, which is the solution with the highest objective function value in the current optimal solution set; otherwise, return to step 5.2.

[0040] Furthermore, the specific method for step 5.3 is as follows:

[0041] Step 5.3.1: Merge the historical best solutions recorded by each algorithm to obtain the current best solution set;

[0042] Step 5.3.2: Prioritize the solutions with the highest objective function values ​​in the current optimal solution set. Each individual constructs the current population;

[0043] Step 5.3.3: Based on the objective function value of the solution and the corresponding algorithm and operator information during the search process of each thread, calculate the contribution of each algorithm and operator to the current population;

[0044] Step 5.3.4: Delete the thread corresponding to the algorithm with the lowest contribution based on the contribution score.

[0045] Furthermore, the method for calculating the contribution of each algorithm and operator to the current population in step 5.3.3 is as follows:

[0046] (7)

[0047] Indicates the first The contribution of each algorithm;

[0048] Indicates the first Algorithms for the current optimal solution set The number of contributing solutions;

[0049]

[0050] To make the current optimal solution set ={ The solution set in the equation is the result of sorting each solution in ascending order of its objective function value. Number of threads; The number of algorithms in the algorithm library; These represent the historical best solution sets of different algorithms;

[0051] Equation (8) represents the... The front of the middle Each solution is selected. ,if The number of solutions in is less than ,but Select all of them .

[0052] Furthermore, the method used in step 3.2 to calculate whether all grids containing the target region are covered by stripes is the ray casting method.

[0053] Furthermore, the aforementioned ray-guiding method refers to:

[0054] If a ray is emitted from a vertex of the grid, and the number of intersections between the ray and the polygon containing the strip is odd, then the vertex is inside the polygon containing the strip.

[0055] Check in turn whether the other vertices of the mesh are inside the polygon where the strip is located. If they are all inside, the mesh is covered by the strip.

[0056] This invention also proposes a multi-satellite mission planning system for large-area targets, comprising the following modules:

[0057] Input module: used to obtain the regional target set And satellite resource set S;

[0058] Strip set calculation module: used to calculate the target set of a region. Each target task strip set i To obtain the regional target set The set of stripes B for all target tasks in the middle, Represents the target set of the region The i-th task, , The number of target tasks in the regional target set;

[0059] Coverage calculation module: used to calculate the regional target set based on the strip set B of all target tasks. Coverage of each target area;

[0060] Planning model building module: used to build a planning model for multi-satellite scheduling tasks in large areas based on the coverage of regional targets;

[0061] Output module: Used to solve the planning model for large-area multi-satellite scheduling tasks and output the planning scheme.

[0062] By adopting the above technical solution, the present invention has the following beneficial effects:

[0063] This invention provides a multi-satellite mission planning method and system for large-area targets. It obtains a set of satellite resources by striping the target area. Then, it grids the target area, finds the set of grids covered by the strips, and calculates the area of ​​the target area covered by the grids. The goal is to maximize the target area covered by the strips and minimize resource usage, thus establishing a planning model. Finally, a depth-adaptive large neighborhood search algorithm is proposed to solve this problem. Due to the simple structure of the planning model, a planning scheme with high coverage observation of regional targets using as few satellite resources as possible can be quickly obtained. Attached Figure Description

[0064] Figure 1 A schematic diagram illustrating a large-area target observed by multiple satellites;

[0065] Figure 2 This is a system flowchart of the present invention;

[0066] Figure 3 A schematic diagram of the large area target coverage calculation process: (a) is a schematic diagram of grid division; (b) is a schematic diagram of using satellite strips to perform task planning on the grid; (c) is a schematic diagram of using the ray method to determine whether the grid has been divided; and (d) is a schematic diagram of the strip-covered area and the uncovered area.

[0067] Figure 4 This is a flowchart of the depth-adaptive large neighborhood search algorithm of the present invention. Detailed Implementation

[0068] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0069] Figures 1 to 4 A specific embodiment of the multi-satellite mission planning method for large-area targets according to the present invention is given, such as... Figure 2 As shown, it includes the following steps:

[0070] Step 1: Obtain the target set of the region And satellite resource set S;

[0071] In this embodiment, the regional target set is preprocessed to filter out the default regional targets.

[0072] Step 2: Calculate the target set for the region Each target task strip set i To obtain the regional target set The set of stripes B for all target tasks in the middle, Represents the target set of the region The i-th task, , This represents the number of target tasks in the regional target set. The specific method is:

[0073] Step 2.1: Based on each target task Based on the location information and observation time information, satellites are selected sequentially. , h =1,2…|S|, where |S| represents the number of satellites participating in the planning within the satellite resource set S. The calculation of satellites... For the target task Observation strip set Finally, the task was obtained. All satellite strip sets , ;

[0074] Step 2.2: Merge all satellite strip sets for all missions to obtain a strip set for large-area targets. .

[0075] In this embodiment, due to the uncertainty of the shape characteristics of large-area targets and the real-time changes of satellite trajectories, and in order to reduce the computational complexity of the algorithm, a satellite strip calculation method based on large-area targets is proposed, which transforms the planning of target tasks into the planning of strips.

[0076] Step 3: Calculate the regional target set based on the strip set B of all target tasks. Coverage of each target area;

[0077] In this embodiment, when solving the multi-satellite mission planning problem for large-area targets, the primary requirement is to meet the coverage requirements of the target mission, followed by minimizing the use of satellite resources while maintaining coverage. Therefore, calculating the coverage of large-area targets is a necessary process, and the quality of the coverage calculation method directly determines the quality of subsequent models and algorithms. Most existing methods first employ a strip parallel segmentation algorithm to divide the large-area target into segments and calculate the coverage of each strip. However, this method can only roughly calculate the coverage of each satellite strip to the regional target strip and has poor stability. Therefore, to accurately calculate the coverage of each large-area target while ensuring the feasibility of subsequent algorithms, this paper first performs gridding on the large-area target, then uses the ray casting method to determine whether each grid is covered, and finally counts all covered grids to calculate the coverage of the large-area target.

[0078] Step 3 calculates the regional target set. The method for determining the coverage of each target area is as follows:

[0079] Step 3.1: Grid the regions containing all targets in the target set R to obtain the coordinates of each grid cell within the target region after grid division. For example... Figure 3 As shown in (a), a large set of targets in a region is gridded. The size of the grid is defined according to the required accuracy, and the grid direction is defined by the latitude and longitude of the earth.

[0080] Step 3.2: Based on the grid's coordinates, determine whether all grids containing the target area are covered by the satellite stripe, and count the number of all grids covered by the satellite stripe. Figure 3 As shown in (b), satellite stripes are used to plan grid tasks.

[0081] In this embodiment, as Figure 3 As shown in (c), the method for determining whether all grids containing the target region are covered by stripes is the ray casting method. The ray casting method is specifically as follows:

[0082] If a ray is emitted from a vertex of the grid, and the number of intersections between the ray and the polygon containing the strip is odd, then the vertex is inside the polygon containing the strip.

[0083] Check in turn whether the other vertices of the mesh are inside the polygon where the strip is located. If they are all inside, the mesh is covered by the strip.

[0084] Step 3.3: Calculate the percentage of the target area in each grid cell to obtain the target area coverage rate, such as... Figure 3 As shown in (d), the area of ​​the covered and uncovered areas is calculated.

[0085]

[0086] The parameter is 0-1, indicating whether the k-th grid is covered by the stripe (1 for yes, 0 for no); m represents the number of grid cells. Indicates the first The area of ​​the target region existing in each grid. This represents the total area of ​​the target region.

[0087] Step 4: Based on the coverage of regional targets, construct a multi-satellite scheduling task planning model for large areas.

[0088] Unlike traditional satellite scheduling and optimization models, this application performs unified optimization across all visible strips of regional targets, employing integrated modeling and abandoning the traditional task planning model with pre-segmented strips. This not only reduces model complexity and eliminates the need to consider time window constraints, but also facilitates global optimization compared to traditional methods. Furthermore, this application designs a reward function based on target switching, providing a more comprehensive consideration of resource utilization and mission requirements.

[0089] Before constructing a mathematical model for large-area multi-satellite scheduling, the following reasonable assumptions are made:

[0090] 1) Given a large set of objectives for a given region, the influence of dynamic or uncertain factors is no longer considered; that is, only the static planning case is discussed.

[0091] 2) Each satellite is an agile satellite, and the data transmission process is no longer considered;

[0092] 3) The location information of each target in the large area is known, and there is no unknown location information or target movement.

[0093] 4) Unexpected situations such as weather or cloud cover are not considered during the planning and scheduling process.

[0094] The task planning model for large-area multi-satellite scheduling is:

[0095] The objective function is:

[0096]

[0097] In equation (1), The primary objective function is to maximize task coverage. It is a secondary objective function, which minimizes the number of stripes used; it only applies if the function... The optimization objective reaches the threshold After that, the function Only then will it be related to the function Simultaneously optimized;

[0098] In equation (2), The parameter is 0-1, indicating whether the k-th grid is covered by the stripe (1 for yes, 0 for no); m represents the number of grid cells. Indicates the first The area of ​​the target region in each grid This represents the total area of ​​the target region.

[0099] In equation (3), Let 'a' be a 0-1 decision variable, representing the first... The first satellite The first transit Whether a strip is used, 1 for yes, 0 for no; This represents the total number of transits within the planning period, where n is the number of satellites participating in the planning. Indicates the first satellite In the A collection of strips of goods that cross the border. Indicates the first satellite In the Total number of stripes per transit.

[0100] The constraints are:

[0101]

[0102] Equation (4) represents the attitude transition relationship constraint between strips during a single transit of the same satellite; Indicates the first The first satellite The first in the strip set of the next transit The end time of each strip; Indicates stripe and strips Conversion time between Indicates stripe and strips The total angle of the conversion between them;

[0103]

[0104] Indicates the first The first satellite The first in the strip set of the next transit Each stripe Indicates the first The first satellite The first in the strip set of the next transit Each stripe Indicates the first The first satellite The first in the strip set of the next transit The start time of each strip; The parameter is 0-1, indicating the first... The first satellite The first in the strip set of the next transit The first stripe and the first All stripes are used, 1 for yes, 0 for no; Represents the i-th satellite's... A collection of strips that cross the border at a time;

[0105] Equation (5) indicates that both stripes must be selected simultaneously for the strip transition time to be considered;

[0106] Equation (6) represents the upper limit constraint on the energy used by each satellite. : indicates satellite The maximum energy available per lap; Indicates satellite The The first in the strip set of the next transit The energy required for each strip.

[0107] Step 5: Solve the planning model for large-area multi-satellite scheduling tasks and output the planning scheme.

[0108] In this embodiment, the method for solving the planning model for large-area multi-satellite scheduling tasks is a deep adaptive large neighborhood search algorithm, such as... Figure 4 As shown, specifically:

[0109] Step 5.1: Randomly generate an initial population, assign each individual in the initial population to a predefined thread as an initial solution, and input the algorithm library and operator library, as well as the algorithm and operator parameters, into each thread.

[0110] In this embodiment, the algorithms in the algorithm library include: 1) hill climbing algorithm; 2) simulated annealing algorithm; 3) tabu search algorithm; 4) overdue acceptance hill climbing algorithm; 5) tabu annealing algorithm; 6) iterative local search algorithm; and a hybrid algorithm composed of the above 5 basic algorithms.

[0111] Step 5.2: Set the algorithm running time. Each thread selects an algorithm from the algorithm library by roulette wheel. At the same time, during the running of each algorithm, operators are selected from the operator library by roulette wheel. Each thread performs a search based on the selected algorithm and operator and records the historical optimal solution of the algorithm. The historical optimal solution refers to the solution with the highest objective function value among the current solution and historical solutions.

[0112] Step 5.3: Merge the historical best solutions of each algorithm to form the current best solution set; retain the solutions with the highest objective function values ​​in the current best solution set. Each individual constructs the current population and calculates the contribution of each algorithm and operator to the current population, and updates the probability of each algorithm and operator being selected based on the contribution.

[0113] The specific method for step 5.3 is as follows:

[0114] Step 5.3.1: Merge the historical best solutions recorded by each algorithm to obtain the current best solution set;

[0115] Step 5.3.2: Prioritize the solutions with the highest objective function values ​​in the current optimal solution set. Each individual constructs the current population;

[0116] Step 5.3.3: Based on the objective function value of the solution and the corresponding algorithm and operator information during the search process of each thread, calculate the contribution of each algorithm and operator to the current population.

[0117] In this embodiment, the contribution is calculated as follows:

[0118] (7)

[0119] Indicates the first The contribution of each algorithm;

[0120] Indicates the first Algorithms for the current optimal solution set The number of contributing solutions;

[0121]

[0122] To make the current optimal solution set ={ The solution set in the equation is the result of sorting each solution in ascending order of its objective function value. Number of threads; The number of algorithms in the algorithm library; These represent the historical best solution sets of different algorithms;

[0123] Equation (8) represents the... The front of the middle Each solution is selected. ,if The number of solutions in is less than ,but Select all of them .

[0124] The method for updating the probability of each algorithm and operator being selected based on contribution is the same as that in the literature "Liu X, LaporteG, Chen Y, et al. An adaptive large neighborhood search metaheuristic for agile satellite scheduling with time-dependent transition time[J]. Computers&Operations Research, 2017, 86: 41-53."

[0125] Step 5.3.4: Delete the thread corresponding to the algorithm with the lowest contribution. In this embodiment, by stopping the execution of the thread corresponding to the algorithm with the worst historical best solution in the thread set, the algorithm's execution speed is accelerated.

[0126] Step 5.4: Randomly pair individuals in the current population into two pairs, run the crossover operator to obtain a new population, and reallocate the initial solution to each thread in a roulette wheel manner according to the objective function value;

[0127] Step 5.5: If the termination condition is met, output the optimal individual, which is the solution with the highest objective function value in the current optimal solution set; otherwise, return to step 5.2.

[0128] The deep adaptive large neighborhood search algorithm in this embodiment is an adaptive algorithm that combines the advantages of local search and evolutionary algorithms. Normally, local search algorithms excel at local optimization but are prone to getting trapped in local optima and lack global optimization capabilities; evolutionary algorithms excel at global optimization but suffer from insufficient convergence and lack local optimization capabilities. The deep adaptive large neighborhood search algorithm combines the strengths of both, exhibiting comprehensive optimization performance. Secondly, the algorithm's adaptability is enhanced through a dual selection process using both an operator library and an algorithm library. During algorithm execution, high-performing algorithms and operators will be used more frequently, while underperforming ones will be gradually phased out, leading to a gradual improvement in the algorithm's overall performance. The simultaneous operation of multiple operators is beneficial for finding the optimal solution. Conventional adaptive large neighborhood search algorithms only design a few common operators, making it difficult to escape local optima during the search process. Therefore, the deep adaptive large neighborhood search algorithm designs multiple operators with different search ranges, complementing each other and improving the algorithm's optimization capabilities. In this embodiment, threads are used as computation modules, and algorithm and operator libraries are used to encapsulate algorithms and operators, facilitating further expansion and improvement. Based on computer configuration and computational requirements, computational threads can be easily added and modified to improve algorithm computing power. Furthermore, when facing different problems with diverse characteristics, new algorithms and operators can be flexibly added to the algorithm and operator library. This embodiment fully considers task timeliness and profitability, and designs a depth-adaptive large neighborhood search algorithm. Internally, a local search algorithm library and operator library are established. Based on problem characteristics, the algorithm adaptively selects neighborhood operators and local search algorithms, effectively solving the problem, reducing the algorithm's solution time, and improving its performance.

[0129] Since this embodiment aims to maximize coverage and minimize the number of strips used to save satellite resources, it utilizes a multi-satellite collaborative observation method for complex observation tasks. This method divides the problem of multi-satellite collaborative observation of large-area targets into two stages: large-area target decomposition and task planning. This solves the observation needs of large-area targets, achieves a coverage rate of over 95% for large-area targets, and meets the user's needs for integrated planning and optimization of large-area targets.

[0130] Unlike traditional calculation methods, a method for dividing large-area targets was established. Large-area targets were discretized into a grid, and the target area within each cell was calculated. The ray-drawing method was used to effectively handle the problem of determining regional coverage. This method breaks through the traditional regional target planning method based on strip pre-segmentation and constructs an integrated model. The approach shifts from a task-oriented to a satellite strip-oriented model, significantly reducing the constraint checking process within the time window and lowering the model's complexity. Simultaneously, target switching and a benefit function that comprehensively considers coverage and resource utilization are adopted. The aim is to reduce satellite resource consumption and improve satellite utilization while ensuring coverage.

[0131] This invention also provides a multi-satellite mission planning system for large-area targets, comprising the following modules:

[0132] Input module: used to obtain the regional target set And satellite resource set S;

[0133] Strip set calculation module: used to calculate the target set of a region. Each target task strip set i To obtain the regional target set The set of stripes B for all target tasks in the middle, Represents the target set of the region The i-th task, , The number of target tasks in the regional target set;

[0134] Coverage calculation module: used to calculate the regional target set based on the strip set B of all target tasks. Coverage of each target area;

[0135] Planning model building module: used to build a planning model for multi-satellite scheduling tasks in large areas based on the coverage of regional targets;

[0136] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A multi-satellite mission planning method for large-area targets, characterized in that, Includes the following steps: Step 1: Obtain the target set of the region And satellite resource set S; Step 2: Calculate the target set for the region Each target task strip set i To obtain the regional target set The set of stripes B for all target tasks in the middle, Represents the target set of the region The i-th task, , The number of target tasks in the regional target set; Step 3: Calculate the regional target set based on the strip set B of all target tasks. Coverage of each target area; Step 4: Based on the coverage of regional targets, construct a large-area multi-satellite scheduling task planning model; the large-area multi-satellite scheduling task planning model is: The objective function is: In equation (1), The primary objective function is to maximize task coverage. It is a secondary objective function, which minimizes the number of stripes used; it only applies if the function... The optimization objective reaches the threshold After that, the function Only then will it be related to the function Simultaneously optimized; In equation (2), The parameter is 0-1, indicating whether the k-th grid is covered by the stripe (1 for yes, 0 for no); m represents the number of grid cells. Indicates the first The area of ​​the target region existing in each grid. This represents the total area of ​​the target region. In equation (3), Let 'a' be a 0-1 decision variable, representing the first... The first satellite The first transit Whether a strip is used, 1 for yes, 0 for no; This represents the total number of transits within the planning period, where n is the number of satellites participating in the planning. Indicates the first satellite In the A collection of strips of goods that cross the border. Indicates the first satellite In the Total number of stripes per transit The constraints are: Equation (4) represents the attitude transition relationship constraint between strips during a single transit of the same satellite; Indicates the first The first satellite The first in the strip set of the next transit The end time of each strip; Indicates stripe and strips Conversion time between Indicates stripe and strips The total angle of the conversion between them; Indicates the first The first satellite The first in the strip set of the next transit Each stripe Indicates the first The first satellite The first in the strip set of the next transit Each stripe Indicates the first The first satellite The first in the strip set of the next transit The start time of each strip; The parameter is 0-1, indicating the first... The first satellite The first in the strip set of the next transit The first stripe and the first All stripes are used, 1 for yes, 0 for no; Represents the i-th satellite's... A collection of strips that cross the border at a time; Equation (5) indicates that both stripes must be selected simultaneously for the strip transition time to be considered; Equation (6) represents the upper limit constraint on the energy used by each satellite. : indicates satellite The maximum energy available per lap; Indicates the first The first satellite The first in the strip set of the next transit Energy required for each band Step 5: Solve the planning model for large-area multi-satellite scheduling tasks and output the planning scheme.

2. The multi-satellite mission planning method according to claim 1, characterized in that, Step 3 calculates the regional target set. The method for determining the coverage of each target area is as follows: Step 3.1: Perform gridding on the regions where all targets are located in the target set R to obtain the coordinates of each grid in the target region after gridding; Step 3.2: Based on the coordinates of the grid, determine whether all grids containing the target area are covered by the stripe, and count the number of all grids covered by the satellite stripe; Step 3.3: Calculate the percentage of the target area in each grid to obtain the target area coverage.

3. The multi-satellite mission planning method according to claim 2, characterized in that, In step 2, the target set of the region is calculated. Each target task strip set i To obtain the regional target set The method for determining the strip set B of all target tasks is: Step 2.1: Based on each task Based on the location information and observation time information, satellites are selected sequentially. , h =1,2…|S|, where |S| represents the number of satellites in the satellite resource set S. The satellites are calculated... For the task Observation strip set Finally, the task was obtained. All satellite strip sets , ； Step 2.2: Merge all satellite strip sets for all missions to obtain a strip set for large-area targets. .

4. The multi-satellite mission planning method according to claim 2, characterized in that, The method for solving the planning model of multi-satellite scheduling tasks for large areas is the deep adaptive large neighborhood search algorithm, specifically: Step 5.1: Randomly generate an initial population, assign each individual in the initial population to a predefined thread as an initial solution, and input the algorithm library and operator library, as well as the algorithm and operator parameters, into each thread; Step 5.2: Set the algorithm running time. Each thread selects an algorithm from the algorithm library by roulette wheel. At the same time, during the running of each algorithm, operators are selected from the operator library by roulette wheel. Each thread performs a search based on the selected algorithm and operator and records the historical optimal solution of the algorithm. The historical optimal solution refers to the solution with the highest objective function value among the current solution and historical solutions. Step 5.3: Merge the historical best solutions of each algorithm to form the current best solution set; retain the solutions with the highest objective function values ​​in the current best solution set. Each individual constructs the current population and calculates the contribution of each algorithm and operator to the current population, and updates the probability of each algorithm and operator being selected based on the contribution. Step 5.4: Randomly pair individuals in the current population into two pairs, run the crossover operator to obtain a new population, and reallocate the initial solution to each thread in a roulette wheel manner according to the objective function value; Step 5.5: If the termination condition is met, output the optimal individual, which is the solution with the highest objective function value in the current optimal solution set; otherwise, return to step 5.

2.

5. The multi-satellite mission planning method according to claim 4, characterized in that, The specific method for step 5.3 is as follows: Step 5.3.1: Merge the historical best solutions recorded by each algorithm to obtain the current best solution set; Step 5.3.2: Prioritize the solutions with the highest objective function values ​​in the current optimal solution set. Each individual constructs the current population; Step 5.3.3: Based on the objective function value of the solution and the corresponding algorithm and operator information during the search process of each thread, calculate the contribution of each algorithm and operator to the current population; Step 5.3.4: Delete the thread corresponding to the algorithm with the lowest contribution based on the contribution score.

6. The multi-satellite mission planning method according to claim 5, characterized in that, The method for calculating the contribution of each algorithm and operator to the current population in step 5.3.3 is as follows: （7） Indicates the first The contribution of each algorithm; Indicates the first Algorithms for the current optimal solution set The number of contributing solutions; To make the current optimal solution set ={ The solution set in the equation is the result of sorting each solution in ascending order of its objective function value. Number of threads; The number of algorithms in the algorithm library; These represent the historical best solution sets of different algorithms; Equation (8) means that The front of the middle Each solution is selected. ,if The number of solutions in is less than ,but Select all of them .

7. The multi-satellite mission planning method according to claim 2, characterized in that, The method used in step 3.2 to calculate whether all grids containing the target region are covered by stripes is the ray casting method.

8. The multi-satellite mission planning method according to claim 7, characterized in that, The aforementioned ray-drawing method refers to: If a ray is emitted from a vertex of the grid, and the number of intersections between the ray and the polygon containing the strip is odd, then the vertex is inside the polygon containing the strip. Check in turn whether the other vertices of the mesh are inside the polygon where the strip is located. If they are all inside, the mesh is covered by the strip.

9. A multi-satellite mission planning system for large-area targets, characterized in that, Includes the following modules: Input module: used to obtain the regional target set And satellite resource set S; Strip set calculation module: used to calculate the target set of a region. Each target task strip set i To obtain the regional target set The set of stripes B for all target tasks in the middle, Represents the target set of the region The i-th task, , The number of target tasks in the regional target set; Coverage calculation module: used to calculate the regional target set based on the strip set B of all target tasks. Coverage of each target area; Planning model construction module: used to construct a planning model for large-area multi-satellite scheduling tasks based on the coverage of regional targets; the planning model for large-area multi-satellite scheduling tasks is: The objective function is: In equation (1), The primary objective function is to maximize task coverage. It is a secondary objective function, which minimizes the number of stripes used; it only applies if the function... The optimization objective reaches the threshold After that, the function Only then will it be related to the function Simultaneously optimized; In equation (2), The parameter is 0-1, indicating whether the k-th grid is covered by the stripe (1 for yes, 0 for no); m represents the number of grid cells. Indicates the first The area of ​​the target region existing in each grid. This represents the total area of ​​the target region. In equation (3), Let 'a' be a 0-1 decision variable, representing the first... The first satellite The first transit Whether a strip is used, 1 for yes, 0 for no; This represents the total number of transits within the planning period, where n is the number of satellites participating in the planning. Indicates the first satellite In the A collection of strips of goods that cross the border. Indicates the first satellite In the Total number of stripes per transit The constraints are: Equation (4) represents the attitude transition relationship constraint between strips during a single transit of the same satellite; Indicates the first The first satellite The first in the strip set of the next transit The end time of each strip; Indicates stripe and strips Conversion time between Indicates stripe and strips The total angle of the conversion between them; Indicates the first The first satellite The first in the strip set of the next transit Each stripe Indicates the first The first satellite The first in the strip set of the next transit Each stripe Indicates the first The first satellite The first in the strip set of the next transit The start time of each strip; The parameter is 0-1, indicating the first... The first satellite The first in the strip set of the next transit The first stripe and the first All stripes are used, 1 for yes, 0 for no; Represents the i-th satellite's... A collection of strips that cross the border at a time; Equation (5) indicates that both stripes must be selected simultaneously for the strip transition time to be considered; Equation (6) represents the upper limit constraint on the energy used by each satellite. : indicates satellite The maximum energy available per lap; Indicates the first The first satellite The first in the strip set of the next transit The energy required for each strip; Output module: Used to solve the planning model for large-area multi-satellite scheduling tasks and output the planning scheme.

Images