A large-scale satellite mission scheduling method, storage medium and device

Abstract

The present invention discloses a large-scale satellite task scheduling method, storage medium and equipment in the field of satellite task allocation technology. The observation conditions of a finite point task set are used to approximate the observation conditions of a regional task; different step sizes are used to perform the initial screening and fine screening processes of the time window to obtain the available time window for the satellite to perform the task; the time window set is encoded, and the algorithm population and the reference point set are initialized; crossover and mutation operations are performed on the chromosomes, and two fitness indicators corresponding to each chromosome are calculated by performing conflict-free projection transformation on the chromosomes; multi-level hierarchical non-dominated sorting operations, adaptive normalization operations, reference point association operations, and microhabitat retention operations are performed on the algorithm population, and the optimal time window is output after repeated iterations. The present invention has a rapid response capability for large-scale satellite task scheduling scenarios, and has a strong adaptive capability for extreme satellite task scheduling scenarios, and can ultimately achieve local optimal matching of satellite task scheduling scenarios.

Description

Technical Field

[0001] The invention relates to a large-scale satellite mission scheduling method, storage medium and equipment, and belongs to the technical field of multi-satellite mission planning and scheduling. Background Art

[0002] Optimization problems involving multiple objective functions are called multi-objective optimization problems; NSGA-III is the third-generation version of the non-dominated sorting genetic algorithm and is one of the most cutting-edge algorithms in the field of multi-objective optimization. Multi-objective optimization is an important technology in the field of task allocation. In engineering practice, multi-objective optimization helps to find Pareto optimal solutions under multiple requirements. With the development of task allocation theory and space scheduling technology, the current satellite scheduling process has the characteristics of precision and coordination, and the task execution process has the characteristics of scale and complexity; and the relevant indicators of the satellite task scheduling process show a trend of strictness and personalization. Therefore, higher requirements are placed on the speed and adaptability of the task allocation process.

[0003] In recent years, the optimal matching problem of large-scale satellite mission scheduling scenarios has received extensive attention, and various optimal matching methods have been proposed. For example, the DRL-based giant constellation satellite-to-ground tracking and control link planning algorithm introduces the deep reinforcement learning method DQN for strategy optimization, but the trained model is difficult to adapt to the situation of changes in the number of satellites; for example, the SAR multi-satellite collaborative complex regional observation planning algorithm based on the improved genetic algorithm uses a complex large-scale regional decomposition algorithm, and adopts an improved genetic algorithm with greedy algorithm initialization, elite retention strategy, and cubic fitness function, but the iterative efficiency and population diversity of the improved genetic algorithm are not guaranteed. Summary of the invention

[0004] The technical problem to be solved by the present invention is to overcome the defects of the prior art and provide a large-scale satellite task scheduling method, storage medium and device.

[0005] In order to solve the above technical problems, the present invention is implemented by adopting the following technical solutions.

[0006] In a first aspect, the present invention provides a large-scale satellite task scheduling method, comprising:

[0007] Step S1: Obtain regional mission raw data, satellite data, and mission data corresponding to each satellite.

[0008] Initialize the available time window and generate the approximate longitude and latitude coordinate sequence of the regional mission using the original data of the regional mission;

[0009] Step S2: traverse each satellite and the tasks corresponding to each satellite, and when the task is within the satellite observation field of view within the preset screening time window, store the start and end time of the screening time window, the deflection angle, and the coverage status to the available time window;

[0010] Step S3: Initialize the parameters of the evolutionary algorithm, let each available time window in step S2 correspond to a gene on the chromosome of the evolutionary algorithm, and generate a priori population, initial population and reference point set;

[0011] Step S4: according to different iterative processes, the population in step S3 or step S7 is used as the parent population, and crossover and mutation operations are performed on each chromosome of the parent population to obtain a progeny population;

[0012] Step S5: for each chromosome in the offspring population of step S4, the time window conflict caused by the overlapping working time of the same satellite, the time window conflict caused by not meeting the task execution interval constraint, the time window conflict caused by not meeting the time constraint required for satellite deflection, and the time window conflict caused by not meeting the satellite continuous operation time constraint are processed to obtain chromosomes with no internal gene conflicts, and the fitness index of each chromosome in the offspring population after the conflict-free projection transformation is calculated;

[0013] Step S6: performing a non-dominated sorting operation on the parent population that has not been subjected to the crossover and mutation operations in step S4 and the offspring population that has been subjected to the conflict-free projection transformation in step S5, and performing an adaptive normalization operation and a reference point association operation on the chromosomes in the sorted population to obtain a sorted population;

[0014] Step S7: performing a niche preservation operation on the sorted population in step S6 to obtain a new generation of population;

[0015] Step S8: After step S7, determine whether the optimal time window has been iterated. If so, convert the optimal fitness value individual in the new generation population obtained in step S7 into a satellite task scheduling plan. Otherwise, return to step S4.

[0016] Furthermore, the generating of the approximate longitude and latitude coordinate sequence of the regional task by using the original data of the regional task includes:

[0017] To the task j The location data class generates longitude and latitude data points within the upper and lower limits of longitude and latitude coordinates ;

[0018] Determine the longitude and latitude data points Is it on task? If it is within the geographical boundary, it is added to the approximate longitude and latitude coordinate sequence In the process of generation and judgment, the number of data points in the approximate longitude and latitude coordinate sequence reaches indivual.

[0019] Furthermore, when the task is within the satellite observation field of view within the preset screening time window, the start and end time of the screening time window, the deflection angle, and the coverage condition are stored in the available time window, including:

[0020] S21, initializing time window screening parameters, including:

[0021] Initialize the time window initial screening iteration step , , time window fine screening iteration step , ,in, is a positive integer, The shortest time required for the satellite to start and end the current mission; initialize the satellite sequence number , Task No. , Available time window number , ; Initialize the initial screening time window and fine-screen time windows , the current filtering time window is simply recorded as ;

[0022] S22, performs a two-stage screening process, including:

[0023] The initial screening process includes:

[0024] Computing Satellites i exist The four vertices of the rectangular field of view corresponding to the time range , , , , here The current screening time window corresponding;

[0025] Get the target point, the target point is Each point in the set, 3D coordinates From spherical coordinates Convert to get, It represents the approximate distance from the sub-satellite point to the center of the star. , n ∈[1, ], Indicates satellite exist The true value of the distance from the sub-star point to the center of the star at this moment, Indicates satellite exist The true value of the distance from the sub-star point to the center of the star at this moment;

[0026] Perform dimensionality reduction operations on the three-dimensional coordinates, and calculate the straight line equations corresponding to the four sides of the rectangular field of view from the four vertices of the rectangular field of view , , by calculating whether the sum of the areas of the triangles formed by the target point and each side of the rectangular field of view is less than the area of ​​the rectangular field of view multiplied by the rectangular field of view judgment threshold. If so, the target point is within the rectangular field of view, otherwise it is not within the rectangular field of view;

[0027] If there is a target point in the rectangular field of view, the task is within the satellite observation field of view within the initial screening time window, and the time window fine screening process is switched; otherwise, the initial screening time window is updated to determine whether the task is within the satellite observation field of view within the new initial screening time window;

[0028] Fine screening process, including:

[0029] After replacing the fine screening time window with the primary screening time window, repeat the primary screening process to determine whether the task is within the satellite observation field of view within the fine screening time window. With the current fine-screening time window Corresponding; if so, calculate the satellite exist Tasks within time frame Coverage and execution of tasks j Required deflection angle, storage of fine-screen time window start and end time, deflection angle, and coverage to satellite i To the task j No. k Available time windows ,make ; Otherwise, update the fine screening time window to determine whether the task is within the satellite observation field of view within the new fine screening time window;

[0030] S23, judging whether the available time windows have been screened, if all satellites and missions have been traversed, then obtaining the available time window set, otherwise updating the traversal object and turning to S22.

[0031] Furthermore, the step S3 comprises:

[0032] S31, generating the initial population, including:

[0033] Initialize the population size according to user needs , prior population size , fitness objective function dimension , Maximum evolutionary generations , Maximum program running time , initialize the number of iterations , program running time ; By satellite To the task No. The start time of the available time window First priority, satellite To the task No. The end execution time of the available time window As the second priority, re-arrange all available time windows in ascending order, and record The gene on the corresponding chromosome , using 0-1 binary encoding, the gene is assigned a value of 0 to represent that the time window is not selected, and the gene is assigned a value of 1 to represent that the time window is selected. The length of each chromosome is ;

[0034] Generate a conflict-free chromosome based on the idea that each satellite performs only one task, and repeat The mutation operation obtains The mutant chromosomes The mutant chromosomes and the original conflict-free chromosomes together constitute a population size of The prior population of

[0035] The chromosome generation process includes: The uniform distribution is sampled, and the sampled values ​​are rounded to get the encoding value of a single gene. The coding value acquisition process of a single gene is repeated to obtain a chromosome; the chromosome generation process is repeated until a chromosome is generated. Chromosomes, the Chromosomes and The prior populations together constitute the initial population ;

[0036] S32, generating a reference point set, including:

[0037] For each fitness objective function The total number of reference points is ,in, is the number of combinations, The value is selected so that the total number of reference points Closest to population size A positive integer value of ;

[0038] From the collection No sequential extraction samples, and get the first set , the first set The size is , the first set The elements in are , the first set middle, ;

[0039] For the first set Perform a mapping operation on each element in to obtain the second set , the second set The elements in are represented as , , the second set middle, ;

[0040] For the second set Perform a mapping operation on each element in to obtain a reference point set , a set of reference points The elements in are represented as , , a set of reference points middle, ,make , , calculate the reference point set ;

[0041] After completing the process of generating the reference point set, the population size Change to the number of reference points.

[0042] Furthermore, the step S5 comprises:

[0043] S51, handling time window conflicts, including:

[0044] 1) Dealing with time window conflicts caused by overlapping working hours of the same satellite, including:

[0045] Initializing the fitness index Part 2 , assuming that the time window conflict caused by the overlapping working time of the same satellite is deducted as , by satellite To the task No. The start time of the available time window Rearrange the time window corresponding to each gene segment of the chromosome in ascending order and compare the same satellite in the chromosome Any two genes , The time window is expressed as:

[0046] ;

[0047] Then , ;

[0048] In the formula, For satellite For any task No. The start time of the available time window, For satellite To the task No. The end execution time of the available time window, For satellite For any task No. The start time of the available time window, For satellite To the task No. The end execution time of the available time window;

[0049] 2) Handling time window conflicts caused by failure to meet task execution interval constraints, including:

[0050] Initializing the fitness index Part 2 , assuming that the single deduction for time window conflict caused by not meeting the task execution interval constraint is ,according to Rearrange the time window corresponding to each gene segment of the chromosome in ascending order and compare the same task in the chromosome Any two genes , The time window is expressed as:

[0051] ;

[0052] Then , ;

[0053] In the formula, For satellite To the task No. The start time of the available time window, For satellite To the task No. The end execution time of the available time window, For the task The minimum time difference between two valid images;

[0054] 3) Handling time window conflicts caused by failure to meet the time constraints required for satellite deflection, including:

[0055] Initializing the fitness index Part 2 , assuming that the single deduction for time window conflict caused by not meeting the task execution interval constraint is ,according to Rearrange the time window corresponding to each gene segment of the chromosome in ascending order and compare the same satellite in the chromosome Next any two , The time window is expressed as:

[0056] ;

[0057] ;

[0058] Then , ;

[0059] In the formula, Indicates satellite From Tasks in available time windows Corresponding deflection angle deflection to Tasks in available time windows The time required for the corresponding deflection angle, For satellite To the task No. The satellite deflection angle for the available time window, Indicates satellite Deflection from zero deflection angle to Tasks in available time windows The time required for the corresponding deflection angle, For satellite To the task No. The satellite deflection angle for the available time window, Indicates satellite Deflection from zero deflection angle to Tasks in available time windows The time required for the corresponding deflection angle;

[0060] 4) Handling time window conflicts caused by failure to meet satellite continuous operation time constraints, including:

[0061] Initializing the fitness index Part 2 Part 4 , assuming that the single deduction for the time window conflict caused by not meeting the satellite continuous operation time constraint is ;

[0062] For each satellite , using satellite The orbital data class statistics the average orbital time of satellites ;according to Rearrange the time window corresponding to each gene segment of the chromosome in ascending order and calculate the same satellite The length of each time window ; For each time window , the statistics satisfy the following formula: The length of the time window sum, if said sum is greater than the satellite The longest cumulative power-on time during one orbit around the star , then let , let the start time The largest time window corresponds to , repeat this process until the time window length The sum is less than or equal to , where Indicates satellite satisfy The start execution time of any available time window of Indicates satellite satisfy The end execution time of any available time window;

[0063] 5) After processing the time window conflict, the chromosome obtained is composed of a chromosome with no internal gene conflict;

[0064] S52, calculates fitness indicators, including:

[0065] After S51 processing, the chromosome with no internal gene conflict is obtained, and the fitness index is calculated by the following formula: :

[0066] ;

[0067] In the formula, is the indicator function, is a norm, It's a task Type, divided into point tasks and area tasks, It's a task The number of times it needs to be executed within the specified time. is the number of data points in the approximate longitude and latitude coordinate sequence, It's a task Importance rating, tasks The actual number of executions and cumulative coverage The calculation formula is as follows:

[0068] ;

[0069] ;

[0070] in, Yes and All-1 vectors of the same dimension, It means to judge each component of the vector separately. Indicates satellite To the task No. Task coverage of available time windows;

[0071] The fitness index 2 is calculated by the following formula ;

[0072] .

[0073] Furthermore, the step S6 comprises:

[0074] S61, the parent population that has not been subjected to the crossover and mutation operations in step S4 and the offspring population that has been subjected to the conflict-free projection transformation in step S5 together constitute the current population, and an efficient non-dominated sorting method is used to sort the current population. The specific steps are as follows:

[0075] 11) The current population is based on the fitness index 1 Sort in ascending order. If the values ​​of the first dimension fitness objective function are the same, sort them according to fitness index 2, and so on to get the arranged population. ;

[0076] 12) Initialize the non-dominated population set ,remember Is the first in the population chromosome, is the total number of chromosomes in the population, represents the first Frontier populations, The larger the population, the higher the priority; initialization ;

[0077] 13) Order represents the total number of current frontier populations, where ,but ,initialization ;

[0078] 14) and All chromosomes in the No The dominant solution of There is no fitness index for all dimensions in And fitness index 2 Both are better than , then Join the Frontier Population If yes, go to step 16); otherwise, go directly to step 15);

[0079] 15) ,if , then let ,Will Join the non-dominated population set , go to 16); otherwise go to 13);

[0080] 16) ,if , the iteration stops and returns the non-dominated population set , otherwise go to 13);

[0081] S62, the set of non-dominated populations returned by S61 The chromosomes in perform adaptive normalization operations, including:

[0082] The S61 population is ,in, Expressing about any Request all The fitness objective function of each chromosome is , , Represents chromosome The multidimensional objective function Dimensional component, , the specific steps are as follows:

[0083] 21) Calculate the ideal point , Represents the first The minimum value of the dimension component,

[0084] ;

[0085] 22) Translate the fitness objective function to obtain the translated fitness objective function , , represents the first Dimensional component,

[0086] ;

[0087] In the formula, represents the ideal point in 21);

[0088] 23) Calculate the extreme points corresponding to each dimension ;

[0089] ;

[0090] ;

[0091] In the formula, represents the extreme point operator, , , represents the first Dimensional component;

[0092] 24) Calculate the intercept , Represents the fitness objective function The intercept component corresponding to dimension,

[0093] Define the hyperplane ,in, for dimension column vector, T To transpose, substitute the extreme points to get:

[0094] ;

[0095] Solve the system of equations and get , the intercept of the hyperplane is , if the above equation has no unique solution, then ;

[0096] 25) Normalize the fitness objective function to obtain the normalized function , normalization function No. j Elements The specific calculation formula is:

[0097] ;

[0098] S63, performing an association operation on the chromosomes in the population after the adaptive normalization operation is performed in S62, including:

[0099] 31) Initialize the reference line vector ;

[0100] ;

[0101] represents the reference point set obtained in S32;

[0102] 32) Calculate each chromosome To each reference line Distance ;

[0103] ;

[0104] 33) Get each chromosome The closest reference line And the corresponding minimum distance ;

[0105] ;

[0106] .

[0107] Furthermore, the step S7 comprises:

[0108] Non-dominated population set The final frontier population is , initialize the retained population set , Indicates The initial population of the round iteration, the number of chromosomes added to the initialization , initialize the reference point set , the specific steps are as follows:

[0109] S71, before calculation The reference line vector in the frontier population The number of chromosomes associated with ;

[0110] ;

[0111] S72, filter the reference line vector set with the least associated chromosomes ;

[0112] ;

[0113] S73, from Randomly select a reference line vector ;

[0114] S74, screening the last frontier population Center and reference line vector Associated chromosome sets ,like , go to S75, otherwise go to S76;

[0115] ;

[0116] S75, retaining chromosomes to the next generation population;

[0117] like ,but , Indicates each chromosome obtained by S63 The closest reference line The corresponding minimum distance, Represents the reference line vector The number of associated chromosomes;

[0118] like , , yes Any chromosome in

[0119] , , , transfer to S77;

[0120] S76, remove reference points , , transfer to S77;

[0121] S77, if , go to S72; otherwise, return the retained population set .

[0122] Furthermore, the step S8 comprises:

[0123] Update program running time , to determine whether the optimal time window has been iterated. If and ,make , return to step S4; otherwise the algorithm terminates.

[0124] In a second aspect, the present invention further provides a computer-readable storage medium storing one or more programs, wherein the one or more programs include instructions, which, when executed by a computing device, cause the computing device to perform the method of the first aspect.

[0125] In a third aspect, the present invention further provides a computer device, comprising:

[0126] One or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the one or more programs include instructions for executing the method of the first aspect.

[0127] The beneficial effects achieved by the present invention are:

[0128] The present invention utilizes an approximate estimation algorithm to decompose regional tasks into a set of point tasks, and can quickly evaluate the coverage of multi-satellite collaborative tasks; adopts different step sizes to perform the initial screening and fine screening processes of time windows, and effectively improves the screening efficiency of available time windows in large-scale mission scenarios; based on the population generation process of conflict-free projection transformation and cross-mutation, it is ensured that the algorithm iteration process still has a certain flexibility while satisfying the constraints of each satellite mission; based on the multi-level hierarchical non-dominated sorting, adaptive normalization, reference point association, microhabitat retention and other operations of two fitness indicators, the solution degradation problem and non-convergence problem that may occur in the algorithm iteration are solved; the present invention has a rapid response capability for large-scale satellite mission scheduling scenarios, and has a strong adaptive capability for extreme satellite mission scheduling scenarios, and can ultimately achieve the local optimal matching of satellite mission scheduling scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0129] Figure 1 It is a flow chart of the two-stage large-scale satellite mission scheduling method based on conflict-free projection transformation and NSGA-III of the present invention;

[0130] Figure 2 This is a comparison chart of the average profit results of the traditional genetic algorithm, the genetic algorithm based on conflict-free projection transformation, and the genetic algorithm based on conflict-free projection transformation and NSGA-III in this example;

[0131] Figure 3 This is a comparison chart of the average running time of the traditional genetic algorithm, the genetic algorithm based on conflict-free projection transformation, and the genetic algorithm based on conflict-free projection transformation and NSGA-III in this example. DETAILED DESCRIPTION

[0132] The present invention will be further described below in conjunction with the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and cannot be used to limit the protection scope of the present invention.

[0133] In the description of the present invention, it should be understood that the terms "center", "longitudinal", "lateral", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first", "second", and the like are only used for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, features defined as "first", "second", and the like may explicitly or implicitly include one or more of the features. In the description of the present invention, unless otherwise specified, "multiple" means two or more.

[0134] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood by specific circumstances.

[0135] Embodiment 1: This embodiment introduces a large-scale satellite task scheduling method, specifically a two-stage large-scale satellite task scheduling method based on conflict-free projection transformation and NSGA-III, comprising the following steps:

[0136] Step S1: Data structure construction stage, obtain satellite data, mission data, initialize the available time window. Use the original data of the regional mission to generate the approximate longitude and latitude coordinate sequence of the regional mission;

[0137] Step S2: Two-stage screening process of available time windows, initializing screening parameters, traversing each satellite and task in step S1, judging whether the task is within the satellite observation field of view in the initial screening and fine screening time windows, and if so, storing the start and end time, deflection angle, and coverage of the fine screening time window to the available time window of step S1;

[0138] Step S3: Generate population and reference point set, initialize the parameters of evolutionary algorithm, let each available time window in step S2 correspond to a gene on the chromosome, generate a priori population, initial population and reference point set;

[0139] Step S4: crossover and mutation stage. According to different iterative processes, the population in step S3 or step S7 is used as the parent population, and crossover and mutation operations are performed on each chromosome of the parent population to obtain the offspring population.

[0140] Step S5: conflict-free projection transformation and fitness index calculation stage, for each chromosome in the offspring population in step 4, the time window conflict caused by the overlapping of the working time of the same satellite, the time window conflict caused by not meeting the task execution interval constraint, the time window conflict caused by not meeting the time constraint required for satellite deflection, and the time window conflict caused by not meeting the satellite continuous operation time constraint are processed, and the fitness index is calculated;

[0141] Step S6: NSGA-III processing stage, applying an efficient non-dominated sorting method named ENS to the parent population that has not performed crossover and mutation operations in step S4 and the offspring population after the conflict-free projection transformation in step S5, and performing adaptive normalization operations and reference point association operations on the chromosomes in the sorted population;

[0142] Step S7: Microhabitat preservation stage, performing microhabitat preservation operation on the population in step S6;

[0143] Step S8: Determine whether the optimal time window iteration is complete.

[0144] The step S1 constructs a data structure according to the satellite model and the mission model, specifically comprising the following steps:

[0145] To the task j The location data class generates longitude and latitude data points according to a certain sampling rule within the upper and lower limits of longitude and latitude coordinates. , for example, with a uniform distribution As a sampling distribution, we have:

[0146] ;

[0147] ;

[0148] ;

[0149] ;

[0150] ;

[0151] ;

[0152] Determine data points Is it on task? j If it is within the geographical boundary, it is added to the approximate longitude and latitude coordinate sequence In the process of generation and judgment, the number of data points in the approximate longitude and latitude coordinate sequence reaches The subsequent calculation and judgment process is replace .

[0153] Here we define some constraint variables and parameters involved, as shown in Table 1:

[0154] Table 1 Constraint variables and parameters

[0155] .

[0156] The step S2 performs a time window primary screening process and a fine screening process, which specifically includes the following steps:

[0157] S21, initializing time window screening parameters;

[0158] Initialize the time window initial screening iteration step , time window fine screening iteration step ,in Is a positive integer. Initialize the satellite number , Task No. , Available time window number Initialize the initial screening time window and fine-screen time windows .

[0159] S22, performing a two-stage screening process;

[0160] 1) Initial screening process:

[0161] Computing Satellites i exist The four vertices of the rectangular field of view corresponding to the time range , , , , here Current screening time window Corresponding.

[0162] Determine whether the target point is within the rectangular field of view. Each point in the set is used as the target point, where the rectangular coordinates From spherical coordinates Convert to get, . Perform dimensionality reduction operations on the three-dimensional coordinates, and calculate the equations of the lines corresponding to the four sides of the rectangular area from the four vertices of the rectangular area , , by calculating whether the sum of the areas of the triangles formed by the target point and each side of the rectangular area is less than the area of ​​the rectangular area multiplied by the rectangular field of view judgment threshold. If so, the target point is within the rectangular field of view; otherwise, it is not within the rectangular field of view.

[0163] If there is a target point in the rectangular field of view, the task is within the satellite observation field of view within the initial screening time window, and the time window fine screening process is switched; otherwise, the initial screening time window is updated to determine whether the task is within the satellite observation field of view within the new initial screening time window.

[0164] 2) Fine screening process:

[0165] Determine whether the task is within the satellite observation field of view within the fine screening time window. The steps are the same as the initial screening process. With the current fine screening time window Corresponding. If in, calculate the satellite i exist Tasks within time frame j Coverage and execution of tasks j Required deflection angle, storage of fine screening time window start and end time, deflection angle, coverage ,make Otherwise, update the fine screening time window to determine whether the task is within the satellite observation field of view within the new fine screening time window.

[0166] The specific calculation process of deflection angle and coverage is as follows: is the iteration start time, is the iteration termination time, the iteration step is 1 second, and the satellite is calculated exist The four vertices of the rectangular field of view corresponding to the time range are Each point in the set is used as the target point, where the rectangular coordinates From spherical coordinates Convert to get, If the target point can be scanned, the scan result is recorded as 1; if the target point cannot be scanned, the scan result is recorded as 0; store the scan results of all target points to get the coverage .

[0167] The set of points that can be scanned is , combined with the cosine theorem to calculate each target point The corresponding deflection angle, each target point corresponds to a deflection angle, and the deflection angle set is obtained Repeat the above process of calculating the deflection angle until all integers in the fine screening time window are traversed, and the deflection angle is calculated as follows:

[0168] ;

[0169] S23, determine whether the available time windows have been screened. If all satellites and missions have been traversed, the available time window set is obtained and the process goes to S3. Otherwise, the traversal object is updated and the process goes to S22.

[0170] Preferably, initializing the algorithm population and the reference point set in step S3 specifically includes the following steps:

[0171] S31, generate the initial population;

[0172] Initialize the population size according to user needs , prior population size , fitness objective function dimension , Maximum evolutionary generations , Maximum program running time , initialize the number of iterations , program running time . Start time First priority, end time As the second priority, re-arrange all time windows in ascending order, and record The gene on the corresponding chromosome Using 0-1 binary coding, the gene is assigned a value of 0 to represent that the time window is not selected, and the gene is assigned a value of 1 to represent that the time window is selected. The length of each chromosome is .

[0173] Generate a conflict-free chromosome based on the idea that each satellite performs only one task, and repeat The mutation operation obtains mutant chromosomes, which together with the original non-conflicting chromosomes constitute a population size of The prior population of .

[0174] from The uniform distribution is sampled, and the sampled values ​​are rounded off to get the encoding value of a single gene. Repeat this process until a chromosome is generated. chromosomes, and this chromosome population together with the prior population constitutes the initial population .

[0175] S32, generating a reference point set;

[0176] A reference point set is generated in a systematic way. Specifically, for each fitness objective function The total number of reference points is ,in is the number of combinations. The value is selected so that the total number of reference points Closest to but probably smaller than the population size A positive integer value from the set No sequential extraction samples, and get the set ,gather The size is ; for the set Perform a mapping operation on each element in , , get the set ; for the set Perform a mapping operation on each element in , , where , , get the reference point set After completing the process of generating the reference point set, the population size Change to the number of reference points.

[0177] In step S4, a crossover and mutation operation is performed on each chromosome of the parent population, and multiple crossover and mutation operations can be compatible, with the simulation of binary crossover operation and polynomial mutation operation as special cases;

[0178] The step S5 calculates the fitness index of each chromosome in the offspring population in step S4, specifically comprising the following steps:

[0179] S51, handling time window conflicts;

[0180] 1) Dealing with time window conflicts caused by overlapping working hours of the same satellite;

[0181] Initializing the fitness index Part 2 , assuming that the time window conflict caused by the overlapping working time of the same satellite is deducted as .according to Rearrange the time window corresponding to each gene segment of the chromosome in ascending order and compare the same satellite in the chromosome Next any two , time window, such as:

[0182] ;

[0183] Then , .

[0184] 2) Handling time window conflicts caused by failure to meet task execution interval constraints;

[0185] Initializing the fitness index Part 2 , assuming that the single deduction for time window conflict caused by not meeting the task execution interval constraint is .according to Rearrange the time window corresponding to each gene segment of the chromosome in ascending order and compare the same task in the chromosome Next any two , time window, such as:

[0186] ;

[0187] Then , .

[0188] 3) Dealing with time window conflicts caused by failure to meet the time constraints required for satellite deflection;

[0189] Initializing the fitness index Part 2 , assuming that the single deduction for time window conflict caused by not meeting the task execution interval constraint is .according to Rearrange the time window corresponding to each gene segment of the chromosome in ascending order and compare the same satellite in the chromosome Next any two

[0190] , time window, such as:

[0191] ;

[0192] in,

[0193] ;

[0194] ;

[0195] Then , .

[0196] 4) Dealing with time window conflicts caused by failure to meet satellite continuous operation time constraints;

[0197] Initializing the fitness index Part 2 Part 4 , assuming that the single deduction for the time window conflict caused by not meeting the satellite continuous operation time constraint is .

[0198] For each satellite First, using satellite The orbital data class statistics the average orbital time of satellites ; Next, press Rearrange the time window corresponding to each gene segment of the chromosome in ascending order and calculate the same satellite The length of each time window ; Then, for each time window , the statistics satisfy the following formula: The length of the time window , if the value is greater than , then let , let the start time The largest time window corresponds to , repeat this process until the time window length The sum is less than or equal to .

[0199] Genes on each chromosome and available time windows There is a one-to-one correspondence, so the chromosome obtained after processing the time window conflict is a chromosome with no internal gene conflict.

[0200] S52, calculate fitness index;

[0201] After S51, the conflict-free chromosome is obtained, and the fitness index 1 is calculated by the following formula:

[0202] ;

[0203] in, is the indicator function, One norm, It's a task Type, divided into point tasks "point" and area tasks "area", It's a task The number of times it needs to be executed within the specified time. is the number of data points in the approximate longitude and latitude coordinate sequence, It's a task Importance rating, tasks The actual number of executions and cumulative coverage The calculation formula is as follows:

[0204] ;

[0205] ;

[0206] in Yes and All-1 vectors of the same dimension, Each component of the vector is judged separately.

[0207] The fitness index 2 is calculated by the following formula:

[0208] ;

[0209] In step S6, a non-dominated sorting operation, an adaptive normalization operation, and a reference point association operation are performed on the parent population in step S4 and the offspring population in step S5, specifically including the following steps:

[0210] S61, performing a multi-level hierarchical non-dominated sorting operation on the S4 parent population and the S5 offspring population;

[0211] The current population is composed of the parent population without crossover and mutation operations in S4 and the offspring population after the conflict-free projection transformation in S5. The current population is sorted using an efficient non-dominated sorting method called ENS. The specific steps are as follows:

[0212] 1) Current population Sort in ascending order according to the first-dimensional fitness objective function value. If the first-dimensional fitness objective function values ​​are the same, sort them according to the second-dimensional fitness objective function value, and so on, to get the arranged population ;

[0213] 2) Initialize the non-dominated population set ,remember Is the first in the population chromosome, is the total number of chromosomes in the population, represents the first Frontier populations, The larger the population, the higher the priority; initialization ;

[0214] 3) Order shows the total number of current frontier populations, where ,but ,initialization ;

[0215] 4) Chromosomes and All chromosomes in the No The dominant solution of There is no fitness objective function value in all dimensions that is better than , then Join the Frontier Population If yes, go to 6); otherwise, go directly to 5);

[0216] 5) ,if , then let ,Will Join the non-dominated population set , go to 6); otherwise go to 3);

[0217] 6) ,if , the iteration stops and returns the non-dominated population set ; Otherwise go to 3).

[0218] S62, performing adaptive normalization operation on chromosomes in the S61 population;

[0219] Adaptive normalization is the normalization of the fitness objective function value of each chromosome to determine the relative superiority of individuals in multi-objective optimization. The S61 population is denoted as , the fitness objective function of each chromosome is , , the specific steps are as follows:

[0220] 1) Calculate the ideal point ;

[0221] .

[0222] 2) Translate the fitness objective function and get ;

[0223] .

[0224] 3) Calculate the extreme points corresponding to each dimension , ;

[0225] ;

[0226] ;

[0227] in ,and .

[0228] 4) Calculate the intercept ;

[0229] Define the hyperplane ,in for Dimensional column vector, substituting into the extreme point, we get:

[0230] ;

[0231] Solving this system of equations, we get , so the intercept of the hyperplane is , if the above equation has no unique solution, then .

[0232] 5) Normalize the fitness objective function and get ;

[0233] .

[0234] S63, performing association operations on chromosomes in population S62;

[0235] In order to ensure that the population evolves towards the Pareto surface of the fitness objective function, it is necessary to build the relationship between the population chromosome and the reference point, that is, to perform an association operation. The specific steps are as follows:

[0236] 1) Initialize the reference line vector ;

[0237] ;

[0238] 2) Calculate each chromosome To each reference line Distance ;

[0239] ;

[0240] in, It's chromosome The fitness objective function of

[0241] 3) Get each chromosome The closest reference line And the corresponding minimum distance ;

[0242] ;

[0243] .

[0244] In step 7, a microhabitat preservation operation is performed on the population in step 6, specifically comprising the following steps:

[0245] Non-dominated population set The final frontier population is , initialize the retained population set , initialize the number of chromosomes added , initialize the reference point set , the specific steps are as follows:

[0246] 1) Before calculation The reference line vector in the frontier population The number of chromosomes associated with ;

[0247] ;

[0248] 2) Filter the reference line vector set with the least associated chromosomes ;

[0249] ;

[0250] 3) From Randomly select a reference line vector ;

[0251] 4) Filter the last frontier population as Center and reference line vector Associated chromosome sets ,like , go to 5), otherwise go to 6);

[0252] 5) Preserve chromosomes to the next generation of population;

[0253] like ,but ;

[0254] like , , yes Any chromosome in

[0255] , , , go to 7);

[0256] 6) Eliminate reference points , , go to 7);

[0257] 7) If , go to 2); otherwise, return the retained population set .

[0258] Preferably, in step 8, it is determined whether the optimal time window has been iterated. If and ,make , return to S4; otherwise the algorithm terminates.

[0259] Embodiment 2, as another embodiment, this embodiment introduces a large-scale satellite task scheduling method, including:

[0260] Depend on Figure 1 As shown, the two-stage large-scale satellite task scheduling method based on conflict-free projection transformation and NSGA-III includes the following specific steps:

[0261] S1, generates the approximate longitude and latitude coordinate sequence of the regional task;

[0262] S2, performs a two-stage screening process for available time windows;

[0263] S21, initializing time window screening parameters;

[0264] S22, performing a two-stage screening process;

[0265] S23, judging whether the available time window has been screened, if all satellites and missions have been traversed, then go to S3, otherwise update the traversal object and then go to S22;

[0266] S3, initialize the algorithm population and reference point set;

[0267] S31, generate the initial population;

[0268] S32, generating a reference point set;

[0269] S4, according to different iterative processes, taking the population in step 3 or step 7 as the parent population, performing crossover and mutation operations on each chromosome of the parent population to obtain the offspring population;

[0270] S5, calculate the fitness index of each chromosome in the S4 offspring population after conflict-free projection transformation;

[0271] S51, handling time window conflicts;

[0272] S52, calculate fitness index;

[0273] S6, performs non-dominated sorting operation, adaptive normalization operation, and reference point association operation on S4 parent population and S5 offspring population;

[0274] S61, perform a multi-level hierarchical non-dominated sorting operation on the S4 parent population and the S5 offspring population

[0275] S62, performing adaptive normalization operation on chromosomes in the S61 population;

[0276] S63, performing association operations on chromosomes in population S62;

[0277] S7, performing niche preservation operations on the S6 population;

[0278] S8, Update program runtime , to determine whether the optimal time window has been iterated. and ,make , return to S4; otherwise the algorithm terminates.

[0279] In this embodiment, a data set is made using two publicly available satellite TLE ephemeris data. Each TLE contains two lines of text, with a total of 69 characters, which can provide sufficient information to determine the orbital position and velocity of the earth-orbiting satellite. These data include the longitude, latitude and altitude of the satellite at each time point and the solar altitude angle of the satellite relative to the subsatellite point. The calculation is completed using SGP4, which is a general and widely accepted standard algorithm suitable for most earth-orbiting satellites. It is mainly used to predict the position and velocity of low-Earth orbit satellites, calculate the orbit of the satellite based on TLE data, and can approximate the orbital perturbations caused by factors such as the earth's gravity, the perturbations of the sun and the moon, and atmospheric drag. The algorithm test scenario is described in JSON format. The scenario file contains the satellites used, key strategic observation points and areas, and possible dynamic adjustment information. The length and structure of this information may vary greatly. The area of ​​the regional mission is in square kilometers. The area is calculated under the EPSG:6933 coordinate reference system. This coordinate system is an equal-area projection suitable for calculating the actual area on the earth. The scene settings are shown in Table 2:

[0280] Table 2 Detailed scene settings

[0281] .

[0282] The simulation experiment was set up with three different scenarios of high, medium and low complexity. Considering the timeliness requirement of multi-satellite mission planning, the simulation experiment set the observation time requirement of one mission to be 24 hours, that is, the simulation time was from 00:00:00 to 23:59:59, and single-day mission planning was performed. Some scheduling planning results of the high-complexity scenario are shown in Table 3:

[0283] Table 3 Results of some planning schemes for high-complexity scenarios

[0284] .

[0285] In order to evaluate the effectiveness and solution effect of the genetic algorithm based on conflict-free projection transformation and NSGA-III (P-NSGA-III) proposed in this paper, a genetic algorithm (GA) and a genetic algorithm based on conflict-free projection transformation (P-GA) were selected for comparison. Each algorithm was performed 5 times in scenarios of the same difficulty. The average algorithm benefit and average running time of the task planning scheme were selected as evaluation criteria.

[0286] The time window screening parameters and genetic algorithm parameter settings used in the simulation are shown in Table 4:

[0287] Table 4 Time window screening parameters and genetic algorithm parameter settings

[0288] .

[0289] like Figure 2 As shown in the figure, the average benefits of the three algorithms in 5 experiments in three scenarios of different complexity are given. It can be seen from the figure that in scenarios of the same complexity, the P-GA algorithm obtains higher benefits than the GA algorithm, and the P-NSGA-III algorithm can further improve the benefits compared with P-GA.

[0290] like Figure 3 As shown in the figure, the average running time results of three algorithms in five experiments in three scenarios of different complexity are given. It can be seen from the figure that in scenarios of the same complexity, the efficiency of the P-GA algorithm is higher than that of the GA algorithm. Compared with P-GA, P-NSGA-III can greatly shorten the running time of the algorithm and effectively improve the running efficiency of the algorithm.

[0291] In summary, the two-stage large-scale satellite mission scheduling method based on conflict-free projection transformation and NSGA-III of the present invention has the characteristics of observation benefits and fast convergence speed. It can effectively solve the problem of large-scale satellite mission scheduling while ensuring that the constraints are not violated as much as possible.

[0292] Embodiment 3 is based on the same inventive concept as other embodiments. This embodiment introduces a computer-readable storage medium storing one or more programs, wherein the one or more programs include instructions, and when the instructions are executed by a computing device, the computing device executes the method described in Embodiment 1.

[0293] Example 4 is based on the same inventive concept as other examples. This example introduces a computer device, including one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the one or more programs include instructions for executing the method described in Example 2.

[0294] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0295] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0296] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0297] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0298] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A large-scale satellite mission scheduling method, characterized in that: include: Step S1: Obtain regional mission raw data, satellite data, and mission data corresponding to each satellite. Initialize the available time window and generate the approximate longitude and latitude coordinate sequence of the regional mission using the original data of the regional mission; Step S2: traverse each satellite and the tasks corresponding to each satellite, and when the task is within the satellite observation field of view within the preset screening time window, store the start and end time of the screening time window, the deflection angle, and the coverage status to the available time window; Step S3: Initialize the parameters of the evolutionary algorithm, let each available time window in step S2 correspond to a gene on the chromosome of the evolutionary algorithm, and generate a priori population, initial population and reference point set; Step S4: according to different iterative processes, the population in step S3 or step S7 is used as the parent population, and crossover and mutation operations are performed on each chromosome of the parent population to obtain a progeny population; Step S5: for each chromosome in the offspring population of step S4, the time window conflict caused by the overlapping working time of the same satellite, the time window conflict caused by not meeting the task execution interval constraint, the time window conflict caused by not meeting the time constraint required for satellite deflection, and the time window conflict caused by not meeting the satellite continuous operation time constraint are processed to obtain chromosomes with no internal gene conflicts, and the fitness index of each chromosome in the offspring population after the conflict-free projection transformation is calculated; Step S6: performing a non-dominated sorting operation on the parent population that has not been subjected to the crossover and mutation operations in step S4 and the offspring population that has been subjected to the conflict-free projection transformation in step S5, and performing an adaptive normalization operation and a reference point association operation on the chromosomes in the sorted population to obtain a sorted population; Step S7: performing a niche preservation operation on the sorted population in step S6 to obtain a new generation of population; Step S8: After step S7, determine whether the optimal time window has been iterated. If so, convert the optimal fitness value individual in the new generation population obtained in step S7 into a satellite task scheduling plan. Otherwise, return to step S4. Generating the initial population includes: Generate a conflict-free chromosome based on the idea that each satellite performs only one task, and repeat The mutation operation obtains The mutant chromosomes The mutant chromosomes and the original conflict-free chromosomes together constitute a population of size The prior population of The chromosome generation process includes: The uniform distribution is sampled, and the sampled values ​​are rounded to get the encoding value of a single gene. The coding value acquisition process of a single gene is repeated to obtain a chromosome; the chromosome generation process is repeated until a chromosome is generated. Chromosomes, the Chromosomes and The prior populations together constitute the initial population .

2. The large-scale satellite task scheduling method according to claim 1, characterized in that: The method of generating an approximate longitude and latitude coordinate sequence of the regional mission using the regional mission original data includes: To the task j The location data class generates longitude and latitude data points within the upper and lower limits of longitude and latitude coordinates ; Determine the longitude and latitude data points Is it on task? If it is within the geographical boundary, it is added to the approximate longitude and latitude coordinate sequence In the process of generation and judgment, the number of data points in the approximate longitude and latitude coordinate sequence reaches indivual.

3. The large-scale satellite task scheduling method according to claim 2, characterized in that: When the task is within the satellite observation field of view within the preset screening time window, storing the start and end time of the screening time window, the deflection angle, and the coverage situation to the available time window includes: S21, initializing time window screening parameters, including: Initialize the time window initial screening iteration step , , time window fine screening iteration step , ,in, is a positive integer, The shortest time required for the satellite to start and end the current mission; initialize the satellite sequence number , Task No. , Available time window number , ; Initialize the initial screening time window and fine-screen time windows , the current filtering time window is simply recorded as ; S22, performs a two-stage screening process, including: The initial screening process includes: Computing Satellites i exist The time range corresponds to the four vertices of the rectangular field of view , , , , here The current screening time window corresponding; Get the target point, the target point is Each point in the set, 3D coordinates From spherical coordinates Convert to get, It represents the approximate distance from the sub-satellite point to the center of the star. , n ∈[1, ], Indicates satellite exist The true value of the distance from the sub-star point to the center of the star at this moment, Indicates satellite exist The true value of the distance from the sub-star point to the center of the star at this moment; Perform dimensionality reduction operations on the three-dimensional coordinates, and calculate the straight line equations corresponding to the four sides of the rectangular field of view from the four vertices of the rectangular field of view , , by calculating whether the sum of the areas of the triangles formed by the target point and each side of the rectangular field of view is less than the area of ​​the rectangular field of view multiplied by the rectangular field of view judgment threshold. If so, the target point is within the rectangular field of view, otherwise it is not within the rectangular field of view; If there is a target point in the rectangular field of view, the task is within the satellite observation field of view within the initial screening time window, and the time window fine screening process is switched; otherwise, the initial screening time window is updated to determine whether the task is within the satellite observation field of view within the new initial screening time window; Fine screening process, including: After replacing the fine screening time window with the primary screening time window, repeat the primary screening process to determine whether the task is within the satellite observation field of view within the fine screening time window. With the current fine-screening time window Corresponding; if so, calculate the satellite exist Tasks within time frame Coverage and execution of tasks j Required deflection angle, storage of fine-screen time window start and end time, deflection angle, and coverage to satellite i To the task j No. k Available time windows ,make ; Otherwise, update the fine screening time window to determine whether the task is within the satellite observation field of view within the new fine screening time window; S23, judging whether the available time windows have been screened, if all satellites and missions have been traversed, then obtaining the available time window set, otherwise updating the traversal object and turning to S22.

4. The large-scale satellite task scheduling method according to claim 3, characterized in that: The step S3 comprises: S31, generating the initial population, including: Initialize the population size according to user needs , prior population size , fitness objective function dimension , Maximum evolutionary generation , Maximum program running time , initialize the number of iterations , program running time ; By satellite To the task No. The start time of the available time window First priority, satellite To the task No. The end execution time of the available time window As the second priority, re-arrange all available time windows in ascending order, and record The gene on the corresponding chromosome , using 0-1 binary encoding, the gene is assigned a value of 0 to represent that the time window is not selected, and the gene is assigned a value of 1 to represent that the time window is selected. The length of each chromosome is ; Generate the initial population ; S32, generating a reference point set, including: For each fitness objective function The total number of reference points is ,in, is the number of combinations, The value is selected so that the total number of reference points Closest to population size A positive integer value of ; From the collection No sequential extraction samples, and get the first set , the first set The size is , the first set The elements in are , the first set middle, ; For the first set Perform a mapping operation on each element in to obtain the second set , the second set The elements in are represented as , , the second set middle, ; For the second set Perform a mapping operation on each element in to obtain a reference point set , a set of reference points The elements in are represented as , , a set of reference points middle, ,make , , calculate the reference point set ; After completing the process of generating the reference point set, the population size Change to the number of reference points.

5. The large-scale satellite task scheduling method according to claim 4, characterized in that: The step S5 comprises: S51, handling time window conflicts, including: 1) Dealing with time window conflicts caused by overlapping working hours of the same satellite, including: Initializing the fitness index Part 2 , assuming that the time window conflict caused by the overlapping working time of the same satellite is deducted as , by satellite To the task No. The start time of the available time window Rearrange the time window corresponding to each gene segment of the chromosome in ascending order and compare the same satellite in the chromosome Any two genes , The time window is expressed as: ； Then , ; In the formula, For satellite For any task No. The start time of the available time window, For satellite To the task No. The end execution time of the available time window, For satellite For any task No. The start time of the available time window, For satellite To the task No. The end execution time of the available time window; 2) Handling time window conflicts caused by failure to meet task execution interval constraints, including: Initializing the fitness index Part 2 , assuming that the single deduction for time window conflict caused by not meeting the task execution interval constraint is ,according to Rearrange the time window corresponding to each gene segment of the chromosome in ascending order and compare the same task in the chromosome Any two genes , The time window is expressed as: ； Then , ; In the formula, For satellite To the task No. The start time of the available time window, For satellite To the task No. The end execution time of the available time window, For the task The minimum time difference between two valid images; 3) Handling time window conflicts caused by failure to meet the time constraints required for satellite deflection, including: Initializing the fitness index Part 2 , assuming that the single deduction for time window conflict caused by not meeting the task execution interval constraint is ,according to Rearrange the time window corresponding to each gene segment of the chromosome in ascending order and compare the same satellite in the chromosome Any two , The time window is expressed as: ； ； Then , ; In the formula, Indicates satellite From Tasks in available time windows Corresponding deflection angle deflection to Tasks in available time windows The time required for the corresponding deflection angle, For satellite To the task No. The satellite deflection angle for the available time window, Indicates satellite Deflection from zero deflection angle to Tasks in available time windows The time required for the corresponding deflection angle, For satellite To the task No. The satellite deflection angle for the available time window, Indicates satellite Deflection from zero deflection angle to Tasks in available time windows The time required for the corresponding deflection angle; 4) Handling time window conflicts caused by failure to meet satellite continuous operation time constraints, including: Initializing the fitness index Part 2 Part 4 , assuming that the single deduction for the time window conflict caused by not meeting the satellite continuous operation time constraint is ; For each satellite , using satellite The orbital data class statistics the average orbital time of satellites ;according to Rearrange the time window corresponding to each gene segment of the chromosome in ascending order and calculate the same satellite The length of each time window ; For each time window , the statistics satisfy the following formula: The length of the time window sum, if said sum is greater than the satellite The longest cumulative power-on time during one orbit around the star , then let , let the start time The largest time window corresponds to , repeat this process until the time window length The sum is less than or equal to , where Indicates satellite satisfy The start execution time of any available time window of Indicates satellite satisfy The end execution time of any available time window; 5) After processing the time window conflict, the chromosome obtained is composed of a chromosome with no internal gene conflict; S52, calculate fitness indicators, including: After S51 processing, the chromosome with no internal gene conflict is obtained, and the fitness index is calculated by the following formula: : ； In the formula, is the indicator function, is a norm, It's a task Type, divided into point tasks and area tasks, It's a task The number of times it needs to be executed within the specified time. is the number of data points in the approximate longitude and latitude coordinate sequence, It's a task Importance rating, tasks The actual number of executions and cumulative coverage The calculation formula is as follows: ； ； in, Yes and All-1 vectors of the same dimension, It means to judge each component of the vector separately. Indicates satellite To the task No. Task coverage of available time windows; The fitness index 2 is calculated by the following formula ; 。 6. The large-scale satellite task scheduling method according to claim 5, characterized in that: The step S6 comprises: S61, the parent population that has not been subjected to the crossover and mutation operations in step S4 and the offspring population that has been subjected to the conflict-free projection transformation in step S5 together constitute the current population, and an efficient non-dominated sorting method is used to sort the current population. The specific steps are as follows: 11) The current population is based on the fitness index 1 Sort in ascending order. If the values ​​of the first dimension fitness objective function are the same, sort them according to fitness index 2, and so on to get the arranged population. ; 12) Initialize the non-dominated population set ,remember Is the first in the population chromosome, is the total number of chromosomes in the population, represents the first Frontier populations, The larger the population, the higher the priority; initialization ; 13) Order represents the total number of current frontier populations, where ,but ,initialization ; 14) and All chromosomes in the No The dominant solution of There is no fitness index for all dimensions in And fitness index 2 Both are better than , then Join the Frontier Population If yes, go to step 16); otherwise, go directly to step 15); 15) ,if , then let ,Will Join the non-dominated population set , go to 16); otherwise go to 13); 16) ,if , the iteration stops and returns the non-dominated population set , otherwise go to 13); S62, the set of non-dominated populations returned by S61 The chromosomes in perform adaptive normalization operations, including: The S61 population is ,in, Expressing about any Request all The fitness objective function of each chromosome is , , Represents chromosome The multidimensional objective function Dimensional component, , the specific steps are as follows: 21) Calculate the ideal point , Represents the first The minimum value of the dimension component, ； 22) Translate the fitness objective function to obtain the translated fitness objective function , , represents the first Dimensional component, ； In the formula, represents the ideal point in 21); 23) Calculate the extreme points corresponding to each dimension ; ； ； In the formula, represents the extreme point operator, , , represents the first Dimensional component; 24) Calculate the intercept , Represents the fitness objective function The intercept component corresponding to dimension, Define the hyperplane ,in, for dimension column vector, T To transpose, substitute the extreme points to get: ； Solve the system of equations and get , the intercept of the hyperplane is , if the above equation has no unique solution, then ; 25) Normalize the fitness objective function to obtain the normalized function , normalization function No. j Elements The specific calculation formula is: ； S63, performing an association operation on the chromosomes in the population after the adaptive normalization operation is performed in S62, including: 31) Initialize the reference line vector ; ； represents the reference point set obtained in S32; 32) Calculate each chromosome To each reference line Distance ; ； 33) Get each chromosome The closest reference line And the corresponding minimum distance ; ； 。 7. The large-scale satellite mission scheduling method according to claim 6, characterized in that: The step S7 comprises: Non-dominated population set The final frontier population is , initialize the retained population set , Indicates The initial population of the round iteration, the number of chromosomes added to the initialization , initialize the reference point set , the specific steps are as follows: S71, before calculation The reference line vector in the frontier population The number of chromosomes associated with ; ； S72, filter the reference line vector set with the least associated chromosomes ; ； S73, from Randomly select a reference line vector ; S74, screening the last frontier population Center and reference line vector Associated chromosome sets ,like , go to S75, otherwise go to S76; ； S75, retaining chromosomes to the next generation population; like ,but , Indicates each chromosome obtained by S63 The closest reference line The corresponding minimum distance, Represents the reference line vector The number of associated chromosomes; like , , yes Any chromosome in , , , transfer to S77; S76, remove reference points , , transfer to S77; S77, if , go to S72; otherwise, return the retained population set .

8. The large-scale satellite mission scheduling method according to claim 7, characterized in that: The step S8 comprises: Update program running time , to determine whether the optimal time window has been iterated. If and ,make , return to step S4; otherwise the algorithm terminates.

9. A computer-readable storage medium storing one or more programs, characterized in that: The one or more programs include instructions which, when executed by a computing device, cause the computing device to perform the method of any one of claims 1 to 8.

10. A computer device, characterized in that: include, One or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the one or more programs include instructions for executing any of the methods described in claims 1 to 8.

Images