An agile remote sensing satellite constellation mission planning method based on improved NSWOA

Abstract

The application discloses an agile remote sensing satellite constellation task planning method based on an improved non-dominated sorting whale optimization algorithm (NSWOA), and relates to the field of satellite constellation task planning.The application considers various constraint conditions in practical engineering, takes maximizing task benefits and minimizing satellite average attitude angles as optimization objectives, and carries out task planning research.The application refers to the updating mechanism of operations such as crossover and mutation in a genetic algorithm, improves the updating mode of individual positions in a traditional NSWOA, and applies the improved NSWOA to agile satellite constellation task planning research.The method adopted by the application is mainly suitable for practical scenes with requirements on task benefits and imaging quality, the improved NSWOA has clear logic, a reasonable process and no need of adjusting parameters, and provides help for agile remote sensing satellite constellation task planning research.

Description

Technical Field

[0001] This invention belongs to the field of satellite mission planning, and in particular relates to an agile remote sensing satellite constellation mission planning method based on an improved NSWOA. Background Technology

[0002] In recent years, aerospace technology has made significant progress, and agile satellites with three-axis rapid maneuverability have been widely used in the field of Earth observation. They not only offer longer observation times but also provide more diverse observation modes. At the same time, as the complexity of observation missions continues to increase, observation methods using a single satellite or a single payload are no longer sufficient to meet the demands. Therefore, collaborative observation methods involving multiple satellites and multiple payloads have become a research focus, and agile satellite constellations composed of optical and SAR payloads are gradually becoming a research hotspot.

[0003] Traditional Earth observation satellite mission planning typically focuses on optimizing a single objective: maximizing observational benefits or completing as many observation tasks as possible. However, this approach neglects the impact of satellite attitude angles on imaging quality and other relevant factors. This can lead to planning schemes that are not optimal in actual implementation and may not achieve the most desirable results in practical applications.

[0004] In view of the shortcomings of the above-mentioned research, this invention proposes an agile remote sensing satellite constellation mission planning method based on an improved multi-objective whale optimization algorithm. Summary of the Invention

[0005] To address the aforementioned shortcomings, this invention proposes an agile remote sensing satellite constellation mission planning method based on an improved multi-objective whale optimization algorithm.

[0006] To achieve the above objectives, the technical solution provided by the present invention is as follows:

[0007] An agile remote sensing satellite constellation mission planning method based on an improved multi-objective whale optimization algorithm includes the following steps:

[0008] Step 1: Given the number of satellites, orbital parameters, and payload type, generate a satellite constellation consisting of optical satellites and SAR satellites; given the latitude and longitude range of the target, generate a given number of large-scale observation imaging point targets and set the priority of the targets;

[0009] Step 2: Based on the constellation, payload, and latitude and longitude of the point targets, perform high-precision orbit recursion to generate the initial observation time window for each satellite for each mission. Delete point targets that cannot be observed by any satellite within the specified time and sort the remaining targets according to the mission number to obtain the observation mission set.

[0010] Step 3: Encode the tasks in the observation task set using the dual-chromosome real number encoding method, and randomly generate m dual chromosomes as the initial population;

[0011] Step 4: Perform constraint detection and conflict resolution on the initial population, and use the resolved population as the first generation population. Calculate the individual target value and increment the iteration count by 1.

[0012] Step 5: Update the individuals in the original population using the improved NSWOA and generate a new advanced population;

[0013] Step 6: Perform constraint detection and conflict resolution on individuals in the high-level population, and calculate the target value;

[0014] Step 7: Merge the original population and the advanced population, and use an elite preservation strategy to generate the final population;

[0015] Step 8: Determine if the current iteration count has reached the maximum iteration count. If so, output the observation task sequence with the highest fitness in Step 7. Otherwise, increment the iteration count by 1, use the current population as the previous generation population, and return to Step 5.

[0016] Furthermore, in step 3, the tasks in the observation task set are encoded using a dual-chromosome real-number encoding method, specifically as follows:

[0017] The first chromosome consists of the observation task sequence, referred to as the X chromosome. x i Let X be the i-th observation task in sequence X, where i = 1, 2, ..., N. i N i The number of observation tasks; the second chromosome consists of satellite serial numbers, referred to as the Y chromosome, which has the same length as the X chromosome and all positions correspond one-to-one. The following method is used to assign values ​​to each position on the Y chromosome: Randomly select an integer from [1, N,] to assign a value to each observation task on the X chromosome; this value is the numerical value at the corresponding position on the Y chromosome. If x... i =j, then it means that the number s is used. j satellite for mission x i Observations are performed, where j∈[1, N] s ], N s This represents the total number of satellites.

[0018] Furthermore, the specific methods for constraint detection and conflict resolution in steps 4 and 6 are as follows:

[0019] Step a: For any observation task sequence The sequence is classified according to the serial number of the observed satellite to obtain satellite s. jObservation task set j = 1, 2, ..., N s N j For satellites s j Number of observation tasks;

[0020] Step b: Traverse satellites s j Observation task set T j Determine T j If the observation tasks in the time window meet the time window constraints, proceed to step c; otherwise, delete the lower priority tasks among the observation tasks that do not meet the requirements and proceed to step c.

[0021] The time window constraint includes the following rules:

[0022] Rule 1: Satellites performing observation missions must have an observation time window for that mission;

[0023] Rule 2: In the same satellite observation task set, the time interval between two adjacent observation tasks must be greater than or equal to the satellite's attitude transition time;

[0024] Step c: Traverse satellites s j Observation task set T j Determine T j Does it meet the satellite resource constraints? If not, then T will be randomly assigned. j The lowest priority task is deleted and satellite resource constraint checks are performed again until T. j The satellite resource constraints must be met, specifically as follows:

[0025] Constraint 1: The number of observation tasks for each satellite cannot exceed the maximum number of observations allowed for the satellite, i.e., N. j ≤N jmax Where j = 1, 2, ..., N s ;

[0026] Constraint 2: The energy consumed by each satellite for imaging and attitude maneuvers must not exceed the satellite's maximum energy limit, i.e. Among them, e oj and e ij Satellites s j Energy consumed for observation and attitude maneuvering, E j For satellites s j The maximum on-board energy capacity.

[0027] Furthermore, the specific method for individual updates using the improved NSWOA described in step 5 is as follows:

[0028] Perform a fast non-dominated sort on all individuals in population P, select the individuals in the first non-dominated front, and add them to set P1. Iterate through all individuals in the population, updating the set for each individual. The specific steps are as follows:

[0029] Step a: Name the i-th individual (currently) as a, and randomly select the j-th individual in the population as b, where j ≠ i. Randomly select an individual c from P1 and perform crossover and mutation operations with individual a to generate a new individual d, and calculate the target value of the new individual.

[0030] Step b: Determine the dominance relationship between individual a and individual d. If individual d dominates individual a, add d to the higher population set P2 and replace individual a with individual d as the new individual a; otherwise, add individual a to P2. Determine the dominance relationship between individual d and individual b. If individual b dominates individual d, add b to the higher population set P2.

[0031] Step c: Calculate position coefficients A and B, and generate a random number p between 0 and 1. Based on the value of p, update part of the position of individual d with position coefficients A or B to generate a new individual d. Perform crossover and mutation operations on the new individual d and individual a. After constraint detection and conflict task resolution, generate the latest individual d and calculate the target value of individual d.

[0032] Step d: Determine the dominance relationship between individual d and individual a. If individual d dominates individual a, then add d to the higher population set P2, replace the original i-th individual, and replace individual a with individual d as the new individual a.

[0033] Step e: Randomly perturb the Y chromosome values ​​at some locations of individual a to generate a new individual e; perform constraint detection and conflict resolution on individual e; and calculate the target value.

[0034] Step f: Randomly select the kth individual from the existing population and name it f, where k ≠ i. Determine the dominance relationship between individual e and individual f. If individual e dominates individual f, then add e to the higher-level population set P2; otherwise, add f to the higher-level population set P2.

[0035] Beneficial effects: The improved NSWOA adopted in this invention has a reasonable process, clear logic, and is easy to implement. Compared with the application of optimization algorithms such as genetics and particle swarm optimization in satellite mission planning, the improved NSWOA has the characteristics of fast convergence speed and less susceptibility to getting trapped in local optima; in addition, compared with traditional single-objective optimization for remote sensing satellite mission planning, this invention considers the impact of imaging quality on the planning results, meeting the actual needs of users. Attached Figure Description

[0036] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below.

[0037] Figure 1 This is a flowchart of the method of the present invention; Detailed Implementation

[0038] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.

[0039] like Figure 1 As shown, the technical solution provided by this invention is an agile remote sensing satellite constellation mission planning method based on an improved NSWOA. Addressing the problems of current traditional remote sensing satellite mission planning methods, this invention proposes a mission planning method based on an improved NSWOA for optimizing objectives such as mission benefits and imaging quality. The specific steps are as follows:

[0040] Step 1, Generate satellite constellation orbits and point targets: Given the number of satellites, orbital parameters, and payload type, generate a satellite constellation consisting of optical satellites and SAR satellites; given the latitude and longitude range of the targets, generate a given number of large-scale observation imaging point targets and set the priority of the targets;

[0041] Step 2, Generate the observation task set: Based on the constellation, payload, and latitude and longitude of the point target, perform high-precision orbit recursion to generate the initial observation time window for each satellite for each task, delete point targets that cannot be observed by any satellite within the specified time, and sort the remaining targets according to the task number to obtain the observation task set.

[0042] Step 3, Generate the initial population: Encode the tasks in the observation task set using a dual-chromosome real number encoding method, and randomly generate dual chromosomes as the initial population;

[0043] This invention employs a dual-chromosome real-number encoding method. The first chromosome consists of the observation task sequence, referred to as the X chromosome. x i Let X be the i-th observation task in sequence X, where i = 1, 2, ..., N. i N i The number of observation tasks; the second chromosome consists of satellite serial numbers, referred to as the Y chromosome, which has the same length as the X chromosome and all positions correspond one-to-one. The following method is used to assign values ​​to each position on the Y chromosome: Randomly select an integer from [1, N,] to assign a value to each observation task on the X chromosome; this value is the numerical value at the corresponding position on the Y chromosome. If x... i =j, then it means that the number s is used. jsatellite for mission x i Observations are performed, where j∈[1, N] s ], N s This represents the total number of satellites.

[0044] Step 4, Constraint Detection and Conflict Resolution, Calculate Target Value: Detect whether there is a conflict between observation tasks, introduce constraint detection and conflict resolution operators to resolve the conflict task, and use the resolved population as the first generation population to calculate the individual target value, and increment the iteration count by 1;

[0045] The specific methods for constraint detection and conflict resolution are as follows:

[0046] Step a: For any observation task sequence The sequences are classified according to the serial numbers of the observed satellites to obtain the observation task sets for each satellite. j = 1, 2, ..., N s N j For satellites s j Number of observation tasks;

[0047] Step b: Traverse satellites s j Observation task set T j Determine T j If the observation tasks in the time window meet the time window constraints, proceed to step c; otherwise, delete the lower priority tasks among the observation tasks that do not meet the requirements and proceed to step c.

[0048] The time window constraint includes the following rules:

[0049] Rule 1: Satellites performing observation missions must have an observation time window for that mission;

[0050] Rule 2: In the same satellite observation task set, the time interval between two adjacent observation tasks must be greater than or equal to the satellite's attitude transition time;

[0051] Step c: Traverse satellites s j Observation task set T j Determine T j Does it meet the satellite resource constraints? If not, then T will be randomly assigned. j The lowest priority task is deleted and satellite resource constraint checks are performed again until T. j The satellite resource constraints must be met, specifically as follows:

[0052] Constraint 1: The number of observation tasks for each satellite cannot exceed the maximum number of observations allowed for the satellite, i.e., N. j ≤N jmax Where j = 1, 2, ..., N s ;

[0053] Constraint 2: The energy consumed by each satellite for imaging and attitude maneuvers must not exceed the satellite's maximum energy limit, i.e. Among them, e oj and e ij Satellites s j Energy consumed for observation and attitude maneuvering, E j For satellites s j The maximum on-board energy capacity.

[0054] The specific steps of constraint detection and conflict resolution operators are as follows:

[0055]

[0056]

[0057] Step 5, Update Individuals and Generate a Higher Population: An improved NSWOA algorithm is used to update individuals and generate a higher population. The method is as follows: Perform a fast non-dominated sort on all individuals in population P, select individuals from the first non-dominated front, and add them to set P1. Iterate through all individuals in the population and update each individual. The specific steps are as follows:

[0058] Step a: Name the i-th individual (the current one) as 'a', and randomly select the j-th individual from the population as 'b', where j ≠ i;

[0059] Step b: Randomly select an individual c from P1 and perform crossover and mutation operations with individual a to generate a new individual d, and calculate the target value of the new individual;

[0060] Step c: Determine the dominance relationship between individual a and individual d. If individual d dominates individual a, then add d to the higher-level population set P2 as the i-th individual of P2, and replace individual a with individual d as the new individual a; otherwise, add individual a to P2 as the i-th individual of P2.

[0061] Step d: Determine the dominance relationship between individual d and individual b. If individual b dominates individual d, then add b to the higher-level population set P2 as the j-th individual in P2.

[0062] Step e: Calculate position coefficients A and B, and generate a random number p between 0 and 1. Based on the value of p, update part of the position of individual d with position coefficients A or B to generate a new individual d.

[0063] Step f: Perform crossover and mutation operations on the new individual d and individual a. After constraint detection and conflict task resolution, generate the latest individual d and calculate the target value of individual d.

[0064] Step g: Determine the dominance relationship between individual d and individual a. If individual d dominates individual a, then add d to the higher population set P2, replace the original i-th individual, and replace individual a with individual d as the new individual a.

[0065] Step h: Randomly perturb the Y chromosome values ​​at some locations of individual a to generate a new individual e. Perform constraint detection and conflict resolution on individual e and calculate the target value.

[0066] Step i: Randomly select the kth individual from the existing population and name it f, where k ≠ i. Determine the dominance relationship between individual e and individual f. If individual e dominates individual f, then add e to the higher-level population set P2 as the kth individual of P2; otherwise, add f to the higher-level population set P2 as the kth individual of P2.

[0067] In step e, the position coefficients A and B need to be calculated separately. The calculation formulas for A and B are: A = numTask / 2 - round(randi(numTask / 2)*Iteration / Max_iteration) + 1, B = round(randi(numTask / 2)*Iteration / Max_iteration) + 1; where numTask is the number of tasks, and Max_iteration and Iteration are the maximum number of iterations and the current number of iterations, respectively.

[0068] Step 6, Constraint Detection and Conflict Resolution, Calculation of Target Value: Perform constraint detection and conflict resolution on individuals in the high-level population, and calculate the individual target value;

[0069] Step 7, Generate a new population: Merge the original population and the advanced population, and use an elite preservation strategy to generate the final population;

[0070] Step 8: Determine if the current iteration count has reached the maximum iteration count. If so, output the observation task sequence with the highest fitness in Step 7. Otherwise, increment the iteration count by 1, use the current population as the previous generation population, and return to Step 5.

[0071] Finally, it should be noted that the above descriptions are merely 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 described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for agile remote sensing satellite constellation mission planning based on improved NSWOA, characterized in that, Includes the following steps: Step 1: Given the number of satellites, orbital parameters, and payload type, generate a satellite constellation consisting of optical satellites and SAR satellites; given the latitude and longitude range of the target, generate a given number of large-scale observation imaging point targets and set the priority of the targets; Step 2: Based on the constellation, payload, and latitude and longitude of the point targets, perform high-precision orbit recursion to generate the initial observation time window for each satellite for each mission. Delete point targets that cannot be observed by any satellite within the specified time and sort the remaining targets according to the mission number to obtain the observation mission set. Step 3: Encode the tasks in the observation task set using the dual-chromosome real number encoding method, and randomly generate m dual chromosomes as the initial population; Step 4: Perform constraint detection and conflict resolution on the initial population, and use the resolved population as the first generation population. Calculate the individual target value and increment the iteration count by 1. Step 5: Update the individuals in the original population using the improved NSWOA algorithm and generate a new higher-level population; perform fast non-dominated sorting on all individuals in population P, select individuals in the first non-dominated front, and add them to set P1; iterate through all individuals in the population and update each individual, with the specific steps as follows: Step a) Name the i-th individual (currently) as a, and randomly select the j-th individual in the population as b, where j ≠ i. Randomly select an individual c from P1 and perform crossover and mutation operations with individual a to generate a new individual d, and calculate the target value of the new individual. Step b): Determine the dominance relationship between individuals a and d. If individual d dominates individual a, add d to the higher population set P2 and replace individual a with individual d as the new individual a; otherwise, add individual a to P2. Determine the dominance relationship between individuals d and b. If individual b dominates individual d, add b to the higher population set P2. Step c) Calculate position coefficients A and B, and generate a random number p between 0 and 1. Based on the value of p, update part of the position of individual d with position coefficients A or B to generate a new individual d. Perform crossover and mutation operations on the new individual d and individual a. After constraint detection and conflict task resolution, generate the latest individual d and calculate the target value of individual d. Step d): Determine the dominance relationship between individual d and individual a. If individual d dominates individual a, add d to the higher population set P2, replace the original i-th individual, and replace individual a with individual d as the new individual a. Step e) Randomly perturb the Y chromosome values ​​at some locations of individual a to generate a new individual e. Perform constraint detection and conflict resolution on individual e and calculate the target value. Step f): Randomly select the kth individual from the existing population and name it f, where k ≠ i. Determine the dominance relationship between individual e and individual f. If individual e dominates individual f, add e to the higher-level population set P2; otherwise, add f to the higher-level population set P2. Step 6: Perform constraint detection and conflict resolution on individuals in the high-level population, and calculate the target value; Step 7: Merge the original population and the advanced population, and use an elite preservation strategy to generate the final population; Step 8: Determine if the current iteration count has reached the maximum iteration count. If so, output the observation task sequence with the highest fitness in Step 7. Otherwise, increment the iteration count by 1, use the current population as the previous generation population, and return to Step 5.

2. The agile remote sensing satellite constellation mission planning method based on improved NSWOA as described in claim 1, characterized in that: In step 3, the tasks in the observation task set are encoded using a dual-chromosome real-number encoding method. The specific implementation method is as follows: The first chromosome consists of the observation task sequence, referred to as the X chromosome. x i Let X be the i-th observation task in sequence X, where i = 1, 2, ..., N. t N t The number of observation tasks; the second chromosome consists of satellite serial numbers, abbreviated as Y chromosome, which is the same length as X chromosome and each position corresponds one-to-one. The following method is used to assign values ​​to each position on the Y chromosome: in [1, N... s Randomly select an integer from the [database name] and assign it a value to the observation task on each X chromosome. This value is the corresponding value on the Y chromosome. If x i =j, then it means that the number s is used. j satellite for mission x i Observations are performed, where j∈[1, N] s ], N s This represents the total number of satellites.

3. The agile remote sensing satellite constellation mission planning method based on improved NSWOA as described in claim 1, characterized in that: The specific methods for constraint detection and conflict resolution described in steps 4 and 6 are as follows: Step a: For any observation task sequence Classify the sequence according to the serial numbers of the observation satellites to obtain the observation task set of satellite s j j = 1, 2,..., N s , where N is the number of tasks observed by satellite s j ​​ Step b: Traverse satellites s j Observation task set T j Determine T j If the observation tasks in the time window meet the constraints, proceed to step c; otherwise, delete the lowest priority task among the observation tasks that do not meet the requirements and proceed to step c. The time window constraint includes the following rules: Rule 1: Satellites performing observation missions must have an observation time window for that mission; Rule 2: In the same satellite observation task set, the time interval between two adjacent observation tasks must be greater than or equal to the satellite's attitude transition time; Step c: Traverse satellites s j Observation task set T j Determine T j Does it meet the satellite resource constraints? If not, then T will be randomly assigned. j The lowest priority task is deleted, and satellite resource constraint checks are performed again until all T tasks are completed. j All meet satellite resource constraints, specifically as follows: Constraint 1: The number of observation tasks for each satellite cannot exceed the maximum number of observations allowed for the satellite, i.e., N. j ≤N jmax Where j = 1, 2, ..., N s ; Constraint 2: The energy consumed by each satellite for imaging and attitude maneuvers must not exceed the satellite's maximum energy limit, i.e. Among them, e oj and e tj Satellites s j Energy consumed for observation and attitude maneuvering, E j For satellites s j The maximum on-board energy capacity.

Images