A Multi-Load Satellite Constellation Design Method Based on the SPBO Algorithm

Abstract

The present invention discloses a design method for a multi-payload satellite constellation based on a student psychological algorithm (SPBO), which relates to the field of satellite constellation design. The present invention takes into account the constraints in the actual engineering of SAR payload satellites and optical payload satellites, uses the revisit time of the emergency mission target area and the average coverage rate as the optimization objectives for the design of the multi-payload satellite constellation, applies SPBO to the design of the multi-payload satellite constellation, and proposes an iterative method for applying this algorithm to the field of satellite constellation design according to the characteristics of SPBO and the characteristics of satellite constellation orbit parameters. The process of the present invention is clear and does not require parameter adjustment, and is easy to implement.

Description

Technical Field

[0001] The present invention belongs to the field of satellite constellation design, and particularly relates to a satellite constellation design method based on the SPBO algorithm. Background Art

[0002] In recent years, with the rapid development of space technology, the traditional single-payload satellite constellation cannot meet the actual requirements of missions, and multiple payloads need to work together. Therefore, multi-payload satellite constellations or hybrid constellations have gradually become the main research direction.

[0003] Since the traditional satellite network cannot meet the growing mission requirements of data services, it is inevitable to increase the number of satellites and the number of constellation orbital planes. Also, because the movement of satellites in orbit follows certain laws in time and space, if multiple satellites are used to complement and connect with each other to complete the same mission, it can not only cover a wider area at the same time but also improve the coverage characteristics of a specific area, thus ensuring that the target area can be covered by satellites at the time interval or coverage multiplicity required by the mission.

[0004] Currently, the algorithms widely used in the field of satellite constellation design are mainly genetic algorithms and particle swarm algorithms. In genetic algorithms, chromosomes share information with each other, and the phenomenon of premature obtaining of the optimal result is likely to occur in the whole population when solving satellite constellation design problems; in particle swarm algorithms, particles only share information through the optimal solution obtained by the current search, and there are defects of local convergence and reduction of population diversity when solving satellite constellation design problems.

[0005] In view of the deficiencies of the above algorithms, the present invention proposes a multi-payload satellite constellation design method based on the SPBO algorithm. Summary of the Invention

[0006] In view of the above defects, the present invention proposes a satellite constellation design method based on the SPBO algorithm for the problem of multi-payload satellite constellation design.

[0007] In order to achieve the above object, the technical solution provided by the present invention is as follows:

[0008] A satellite constellation design method based on the SPBO algorithm, comprising the following steps:

[0009] Step 1, under the constraint conditions of a given target area and the number of constellation satellites, determine the range of initial variables, and randomly generate a certain number of satellite orbital parameters, including the satellite orbital height H n , the argument of perigee of the SAR payload satellite , the argument of perigee of the optical payload satellite , the orbital inclination of the SAR payload satellite , and the orbital inclination of the optical payload satellite

[0010] Step 2: By interconnecting with the Satellite Tool Kit (STK), obtain the maximum revisit interval time and average coverage rate of a given target area. Generate the revisit time data Z_max of the satellite constellations determined by different orbital elements for the target area in this population. Combine with the orbital parameter data generated in Step 1 as the initial population of the SPBO algorithm. The expression of any population individual is: T = {T1, T2, …, T n , …, T N}, where, T n represents the individual in the T-th population, and T n contains H n , Z_max n , n = 1, 2, …, N;

[0011] Step 3: Arrange the initial population T in ascending order according to the size of the maximum revisit time interval, and use the sorted population as the population obtained from the first iterative calculation.

[0012] Step 4: Update the previous generation population based on the SPBO algorithm to obtain a new population. For the population after the algorithm update operation, calculate the maximum revisit time interval of each population individual according to the orbital parameters in the new population. Arrange the calculated new population in ascending order according to the size of the revisit time interval. The sorted new population is used as the population obtained from the current iterative calculation, and retain the satellite orbital parameters and revisit time data corresponding to the minimum maximum revisit time interval.

[0013] Step 5: Determine whether the current iteration number is greater than the maximum iteration number. If so, output the individual with the largest fitness value in Step 4; otherwise, add 1 to the iteration number, use the population obtained from the current iterative calculation as the previous generation population, and return to Step 4.

[0014] Furthermore, the specific implementation method of generating the initial population in Step 1 is as follows: According to the number of satellites in the proposed constellation being 2n, generate three initial parameter matrices for the initial population pop: the orbital altitude matrix H N , the mean anomaly and argument of perigee parameter matrices and the SAR orbital inclination matrix and the optical payload satellite orbital inclination Limit the orbital altitude range to H min ~H max , the argument of perigee and mean anomaly limit range is 0° to 360°, the SAR payload satellite orbital inclination limit range is 0° to 180°, and the optical payload satellite orbital inclination limit range is 90° to 100°. Since the mean anomaly and argument of perigee ranges are the same, the two variables are generated together. With The matrix dimension is n*(2pop), where WO w ={WO1, WO2, …, WO pop} is the mean anomaly, and WO o ={WO pop+1 , WO pop+2 , …, WO (2pop)} is the argument of perigee parameter.

[0015] Furthermore, the specific implementation methods for calculating the fitness values in steps 2 and 4 are as follows: Grid the task target area, mainly including the longitude and latitude information of the target area boundary, the regional sampling resolution GP rate, etc. First, based on the known longitude and latitude information of the target area boundary, outline the specific shape and scope of the target area; then divide the target area according to the regional sampling resolution, and take each point of the grid as a data point P N Transmit and save it, where P N ={P1, P2, …, P n}, N represents the number of data points in the target area, and P n represents the nth point area; the smaller the regional sampling resolution GP (generally, GP is taken as 0.5), the larger N is.

[0016] Furthermore, the revisit time interval is opposite to the concept of the fitness value. The revisit time interval is the time difference between two visits to a certain position point in the target area under this satellite constellation. The larger the revisit time interval, the smaller the fitness value.

[0017] Since the constellation set in step 1 will pass through the data points multiple times within the time of this simulation scenario, so by recording the time of the position where this data point is visited once, record the access time when the satellite passes through this data point the next time. According to the time difference n between the two times of the same data point, the revisit interval Time

[0018] Furthermore, step 4 retains the satellite orbit parameters and revisit time data corresponding to the minimum of the maximum revisit time intervals. The specific implementation method is as follows: After calculating the revisit interval Time n for the constellations in the current population, select the maximum value from the N revisit times as the maximum revisit time of this constellation. Repeat this step pop times, then retain the minimum value in Time n , and take this value as the optimal individual of this generation of population. This individual will be compared with the optimal individuals generated in subsequent populations.

[0019] Further, the specific operation of updating the sorted population based on the SPBO algorithm in step 4 is as follows: The satellite constellation orbit parameters with the top k in fitness value and revisit time interval in the population are regarded as the most excellent students, the satellite constellation orbit parameters ranked k + 1 to d are regarded as excellent students, the satellite constellation orbit parameters ranked d + 1 to y are regarded as ordinary students, and the satellite constellation orbit parameters ranked y + 1 to pop are regarded as the last students; k < d < y < pop and k, d, y, pop are all positive integers.

[0020] Further, the single-point mutation operation and crossover operation are used to adjust the most excellent students; the single-point mutation and adjacent swap operation are used to adjust the excellent students, and a part of the excellent students are selected from the adjusted excellent students, and the randomly selected excellent students are adjusted again by the crossover operation; the adjacent swap mutation operation and crossover operation are used to adjust the ordinary students; the single-point mutation and crossover operation are used to adjust the last students.

[0021] Further, the specific implementation method of population update in step 4 is as follows: For an individual X in the new population obtained after crossover and mutation using the SPBO algorithm in step 4 i . X i 's fitness is RT(X i ), and the performance of this population in the previous generation population is X i-1 , X i-1 's fitness value is RT(X i-1 ), and X i is updated using the following formula:

[0022]

[0023] Among them, X represents the revisit time interval of the updated satellite constellation to the target area, rand represents a random number between 0 and 1, and a is the threshold value.

[0024] Beneficial effects:

[0025] The SPBO algorithm adopted by the present invention has a simple concept, clear process and does not require parameter adjustment, and is easy to implement. Compared with the application of genetic algorithm and particle swarm algorithm in satellite constellation design, the SPBO algorithm adopted by the present invention has the characteristics of fast convergence speed and is not easy to fall into local optimum in the later stage of iteration. Brief Description of the Drawings

[0026] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required to be used in the embodiments will be briefly introduced below.

[0027] Figure 1 It is the flowchart of the embodiment of the present invention. Detailed Embodiment

[0028] Next, in combination with the accompanying drawings in the embodiments of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0029] To make the purpose, technical solution and advantages of the present invention clearer and more understandable, the following further details the present invention in combination with specific embodiments.

[0030] As Figure 1 shown, the technical solution provided by the present invention is a multi-payload satellite constellation design method based on the SPBO algorithm. For emergency tasks, the constellation satellites receive requests for earth observation tasks and conduct multiple visits and complete imaging of the target area within a certain period of time. The target area to be covered is disassembled according to the task requirements, and at the same time, based on the existing number of constellation satellites, the initial orbit parameters are generated based on the sun-synchronous regression orbit parameters. Subsequently, the maximum revisit time interval and coverage rate of the current satellite constellation for the position points in the target area are calculated and it is judged whether the iteration number limit is reached. If the iteration number limit is not reached, the operation continues. If the upper limit of the number is reached, the optimal individual in the current population is output, and this individual includes the orbit parameters of the satellite constellation, the maximum revisit time interval, and the coverage rate of the satellite constellation for the target area. The specific steps are as follows:

[0031] Step 1, generate the satellite constellation orbit: Under the constraint conditions of a given target area and the number of constellation satellites, a given initial variable range is set, and a certain number of satellite orbit parameters are randomly generated. The parameters include altitude h N , argument of perigee w N , orbital inclination i N , etc.

[0032] The M regional targets are meshed and divided into m position points. Based on the planned tasks, the number of satellites in the constellation is designed to be 2n, the number of target areas is M, and the initial population number is pop. On this basis, three initial parameter matrices are generated: the orbital altitude matrix H N , the matrix of mean anomaly at perigee and argument of perigee and the orbital inclination matrix of SAR payload satellites and the orbital inclination matrix of optical payload satellites At the same time, the orbital altitude range H min ~H max, the argument of perigee and mean anomaly are limited to the range of 0° to 360°, the SAR orbit inclination is limited to the range of 0° to 180°, and the optical payload satellite orbit inclination is limited to the range of 90° to 100°. Since the mean anomaly and argument of perigee have the same range, the two variables are generated together. In the generated matrix, H N has a dimension of 2n*pop, has a dimension of n*pop, has a dimension of n*pop, and the matrix has a dimension of n*(2pop), where WO w ={WO1, WO2, …, WO pop} is the mean anomaly, and WO o ={WO pop+1 , WO pop+2 , …, WO (2pop)} is the argument of perigee parameter.

[0033] Based on the above initial matrix, a satellite constellation is built:

[0034]

[0035]

[0036] The initial orbit parameter matrix is shown in Table 1 below:

[0037] Table 1 Orbit Parameters - pop*2n Example Table of Mean Anomaly and Argument of Perigee Parameter Matrix

[0038]

[0039] In the table, columns 1 to pop are the mean anomaly parameters, and columns pop + 1 to 2pop are the argument of perigee parameters.

[0040] Step 2: Obtain the maximum revisit interval time and average coverage rate of a given target area by interconnecting with STK, and obtain the revisit time data of the satellite constellation as the objective function value of the initial population.

[0041] First, build a satellite constellation according to the orbit parameters generated in Step 1, and generate an initial population based on the revisit interval time obtained from the report of the satellite constellation.

[0042] In view of the emergency mission background, the present invention uses the revisit interval time as the constellation optimization index. The revisit interval time is expressed as the interval time between the data points in Step 1 and satellite communication. The shorter this time is, the shorter the revisit period is, and the more times the satellite revisits the target area. It is an important criterion for judging whether a satellite constellation can perform an emergency mission.

[0043] The specific steps for calculating the revisit time interval of the satellite constellation are as follows:

[0044]

[0045]

[0046] Through the above steps, the maximum revisit time interval data Z_max of the satellite constellation determined by different orbital elements in the population for the target area is obtained, and Z_max is used as the objective function value of the individuals in the initial population. The expression of any population is: T = {T1, T2, …, T n , …, T N}. Among them, T n represents the nth individual in population T,

[0047] Step 3: Sort each individual in the previous generation population in ascending order according to the revisit time interval and retain the constellation orbital parameters with the smallest maximum revisit time interval.

[0048] Specifically, through Step 2, the revisit interval Time n of the constellation in the current population is calculated. Select the maximum value from the N revisit times as the maximum revisit time interval MaxRTime of the constellation. After repeating the above steps pop times, retain the revisit interval Time n to form the maximum revisit time sequence of the population and select the minimum value (i.e., the one with the largest fitness value and the best fitness) for retention. This value is the optimal individual of this generation of the population, and this individual will be compared with the optimal individuals generated in subsequent populations.

[0049] Among them, the revisit time interval is the time difference between two consecutive visits to a certain position point in the target area under the satellite constellation. The smaller the maximum revisit time interval, the larger the fitness value and the better the fitness. The calculation method of the revisit time interval / fitness value is as follows:

[0050] First, perform grid processing on the target area, including the longitude and latitude information of the target area boundary and the regional sampling resolution GP rate; according to the known longitude and latitude information of the target area boundary, draw the specific shape and range of the target area, and then divide the target area according to the regional sampling resolution. Take each point of the grid as the data point P N and transmit and save it, where P N = {P1, P2, …, P n}, N represents the number of data points in the target area, and P n represents the nth point area; the smaller the regional sampling resolution GP (generally GP takes 0.5), the larger N is.

[0051] Since the constellation set in Step 1 will pass through the data points multiple times within the simulation scenario time, the time at the location of the data point visited this time is recorded once. When the satellite passes through this data point next time, record the access time. According to the time difference between the two times of the same data point the revisit interval Time of this location under the current constellation can be determined. n .

[0052] Step 4: Classify the satellite constellation orbit parameters in the previous generation population and the maximum revisit time of the constellation constructed from these parameters into four categories according to students' psychology: the best, excellent, ordinary, and end students, and adopt different iteration methods for individuals of different categories to form a new population.

[0053] The SPBO formula proves that this algorithm is applicable to solving numerical optimization problems in the real number domain. However, it has not been applied to satellite constellation design. In the present invention, the satellite constellation uses the data of the maximum revisit time. According to the characteristics of the SPBO algorithm and the characteristics of satellite constellation orbit parameters, an iteration method for applying this algorithm to the field of satellite constellation design is proposed.

[0054] The multi-payload satellite constellation design based on SPBO consists of the following 4 iterative stages:

[0055] Regarding the satellite constellation orbit parameters with the top k fitness values and revisit time intervals in the population as the best students, those with the rankings of k + 1 to d as excellent students, those with the rankings of d + 1 to y as ordinary students, and those with the rankings of y + 1 to pop as end students; k < d < y < pop and k, d, y, pop are all positive integers.

[0056] Next, the definition and mutation crossover operations for the four student categories will be carried out.

[0057] For the best students with the best fitness orbital parameters, single-point mutation operation and crossover operation are used to adjust the best students.

[0058] The excellent students are adjusted by single-point mutation and adjacent swap mutation operations. Among the adjusted excellent students, a part of the excellent students are selected, and crossover operation is used to adjust the randomly selected excellent students again.

[0059] The ordinary students correspond to the task sequences with medium and lower fitness rankings. Adjacent swap mutation operation and crossover operation are taken to adjust these ordinary students.

[0060] The end students correspond to the task sequences with the lowest fitness rankings. Single-point mutation operation and crossover operation are taken for these task sequences to adjust.

[0061] The specific operations are as follows:

[0062]

[0063]

[0064]

[0065] In the above program, according to the sorted population, the samples are divided into four parts, and an individual in the population is randomly selected and judged. If the individual is the one with the highest fitness, the orbital parameters in this individual are randomly assigned to one of the remaining individuals in the group to complete the update operation. If the individual belongs to the group with above-average fitness, the parameters of the individual with the highest fitness are assigned to this individual; if the individual belongs to the group with below-average or lower fitness, this individual is cross-exchanged with its adjacent individual to complete the update operation.

[0066] For the population updated by the SPBO operation in step 4, the newly generated satellite orbit data is imported into STK, and the fitness value of each individual in the population is calculated according to the fitness value calculation method mentioned in step 2.

[0067] For an individual X in the new population obtained after the SPBO crossover and mutation in step 4 i re-sort the fitness value / maximum revisit time interval, and compare it with the individuals in the population before mutation. If the fitness of the new individual is good, keep it; otherwise, randomly keep the new individual or keep the individual before mutation. Specifically:

[0068] Traverse the population, for an individual X in the population after mutation i , its fitness is RT(X i ), the performance of this population in the previous generation population is X i-1 , the fitness value of X i-1 is RT(X i-1 ), and the following formula is used to update X i :

[0069]

[0070] Among them, X represents the revisit time interval of the updated satellite constellation to the target area, rand represents a random number between 0 and 1, and a is the threshold value.

[0071] Step 5, determine whether the current iteration number is greater than the maximum iteration number. If so, output the individual with the largest fitness value in step 4; otherwise, add 1 to the iteration number, use the population calculated in the current iteration as the previous generation population, and go to step 4.

[0072] In addition, it should be noted that, among the various specific technical features described in the above specific embodiments, without conflict, they can be combined in any appropriate manner. To avoid unnecessary repetition, the present invention will not separately describe various possible combination manners.

[0073] Finally, it should be noted that the above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A multi-payload satellite constellation design method based on the SPBO algorithm, characterized in that, It includes the following steps: Step 1, under the constraints of a given target area and the number of constellation satellites, randomly generate a certain number of satellite orbital parameters and build an initial satellite constellation. Assume the number of satellites in the constellation is 2n, and the initial population size is pop. On this basis, generate three initial parameter matrices: the orbital altitude matrix H N , the matrix of mean anomaly at perigee and argument of perigee and the matrix of SAR orbital inclination and the orbital inclination of optical payload satellites and and The matrix dimension is n*(2pop), where WO w ={WO1, WO2, …, WO pop} is the mean anomaly at perigee, and WO o ={WO pop+1 , WO pop+2 , …, WO (2pop)} is the argument of perigee parameter; Step 2: Interconnect with the STK software to obtain the maximum revisit time interval and average coverage rate of the satellite constellation for the specified target area, and generate an initial population based on the satellite orbit parameters. The method for generating the initial population is as follows: Generate the initial population according to several satellite constellation orbit parameters, and use the maximum revisit time interval data Z_max for the target area as the objective function value of the individuals in the initial population. The initial population, combined with the satellite orbit parameter data generated in Step 1, serves as the initial population of the SPBO algorithm. Any population is expressed as T = {T1, T2, …, T n , …, T N}, and each individual in the population is expressed as T n , where Step 3: Calculate the maximum revisit time interval of the initial population, arrange the population individuals in ascending order according to the size of the maximum revisit time interval, and use the arranged population as the first-generation population; The revisit time interval is the time difference between two visits to a certain position point in the target area under the satellite constellation, and the calculation method is as follows: Perform grid processing on the target area, divide the target area according to the regional sampling resolution, and use each point of the grid as the data point P N Transmit and save; since the satellite constellation in step 1 will pass by the data point multiple times within the simulation scenario time, by recording the time at the location of the currently visited data point Make a record once, and record the access time when the satellite passes by the data point next time According to the time difference between the previous and next times of the same data point Determine the revisit interval Time of this location under the current constellation n ; Step 4: Update the previous-generation population based on the SPBO algorithm to obtain a new population, sort and update the new population individuals in ascending order according to the size of the maximum revisit time interval, use the updated population as the population obtained by the current iterative calculation, and retain the satellite orbit parameters and revisit time interval data corresponding to the minimum maximum revisit time interval; The SPBO algorithm is as follows: Divide the satellite constellation orbit parameters and the corresponding maximum revisit time / fitness of the satellite constellation for the target area in the previous-generation population into four categories: the most excellent, excellent, ordinary, and end students according to the students' psychology, and adopt different iterative methods for individuals of different categories to form a new population; among them, The most excellent student corresponds to the orbit parameters with the minimum maximum revisit time / best fitness, and single-point mutation operation and crossover operation are used to adjust the most excellent student; The excellent students are adjusted by single-point mutation and adjacent swap mutation operations. Select a part of the excellent students from the adjusted excellent students and use the crossover operation to adjust the randomly selected excellent students again; The ordinary students correspond to the task sequences with backward fitness rankings, and are adjusted by adjacent swap mutation operations and crossover operations; The end students correspond to the task sequences with the last fitness rankings, and are adjusted by single-point mutation operations and crossover operations; The method for generating the new population is as follows: Randomly select an individual from the population and make a judgment. If the individual belongs to the most excellent student, randomly assign the orbit parameters in the individual to one of the other individuals in the population to complete the update operation; if the individual belongs to the excellent student, assign the individual parameters with the highest fitness to the individual; if the individual belongs to the group below the middle and below, cross-swap the individual with its adjacent individual to complete the update operation; Step 5: Judge whether the current iteration number is greater than the maximum iteration number. If so, output the individual with the minimum maximum revisit time interval in Step 4; otherwise, add 1 to the iteration number, use the population obtained by the current iterative calculation as the previous-generation population, and return to Step 4.

2. The multi-payload satellite constellation design method based on the SPBO algorithm according to claim 1, characterized in that, The operation of sorting and updating individuals in the new population in step 4 is as follows: Calculate an individual X in the new population obtained after the update operation of the SPBO algorithm i , X i 's fitness is RT(X i ). The performance of this population in the previous generation population is X i-1 , X i-1 's fitness value is RT(X i-1 ). Use the following formula to update X i : Among them, X represents the revisit time interval of the updated satellite constellation for the target area, rand represents a random number between 0 and 1, and a is the threshold value.

3. The multi-payload satellite constellation design method based on the SPBO algorithm according to claim 1 or 2, characterized in that, The population retention operation in Step 4 is as follows: Based on the maximum revisit time interval MaxRTime of the constellation in the current population, repeat pop times to obtain the maximum revisit time sequence of the population, select the minimum value for retention, and the individual corresponding to the minimum value is the optimal individual of the current generation population.

Images