GA-SA hybrid optimization processing method for earth observation satellite, storage medium and system

Abstract

The invention relates to the technical field of satellite observation scheduling, in particular to a GA-SA hybrid optimization processing method for earth observation satellites, a storage medium and a system.The hybrid optimization processing method is of a two-layer parallel search structure, and group parallel optimization is carried out, specifically, a group is initialized, and GA and SA search is carried out in sequence at all temperatures, an initial solution of the SA comes from an evolution result of the GA, a solution obtained by the SA through a Metropolis sampling process is used as an initial population of the next round of evolution of the GA, and loop iteration is carried out until convergence is achieved. Compared with the prior art, the method solves the problems that most of single solution algorithms in the prior art are prone to premature and fall into local optimum, the optimization performance of the algorithm can be improved, and the optimization performance and robustness of the satellite observation task planning problem are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of satellite observation scheduling, and in particular to a GA-SA hybrid optimization processing method, storage medium and system for earth observation satellites. Background Art

[0002] Earth observation satellites have become an important tool for scientific research, meteorological remote sensing, agriculture, forestry, emergency search and rescue, etc. With the increase in the number of satellites and the continuous upgrading of application requirements, higher requirements are placed on satellite resource scheduling. Satellite mission planning is a goal optimization problem that meets constraints. Under the condition of limited resources, time and space, reasonable mission planning can improve the efficiency of satellite work.

[0003] The mission planning problem is a complex combinatorial optimization problem. When the types of satellites are different or the types of tasks to be performed are different, it is necessary to establish a planning model and solution algorithm corresponding to the type of satellite or task. Different types of satellites have different working modes, which leads to different specific constraints in mission planning and different goals of mission planning. The differences in constraints and optimization goals make the corresponding planning models and solution algorithms not universal, so various types of mission planning models and solution algorithms have been generated. In recent years, the rapid development of microsatellites has led to a sharp increase in the number of satellites in space. These satellites often conduct networked formation observations. Users have increasingly higher requirements for the diversification, complexity and information fusion of observation tasks. Emergency tasks are required in wartime and natural disasters. These situations have greatly increased the load on ground measurement and control stations, making it difficult to perform intelligent management and control for complex multi-tasks.

[0004] Compared with the mission planning of traditional satellites, the mission planning of Earth observation satellites faces more complexity and constraints. First, Earth observation satellites need to obtain high-quality data within a specific time and space window, and the mission planning problem needs to consider the time window of satellite movement. In addition, Earth observation satellites also face the challenges of data transmission and storage. They must scientifically arrange data download and processing under limited bandwidth and storage capacity to ensure the efficiency and effectiveness of mission execution. Therefore, the mission planning of Earth observation satellites requires not only more efficient algorithms and models, but also in-depth analysis of the complex interactions between the external environment and internal systems.

[0005] However, most of the current single solution algorithms are prone to premature maturity, falling into local optimality, and the degree of improvement is limited. Summary of the invention

[0006] The purpose of the present invention is to solve at least one of the above problems and to provide a GA-SA hybrid optimization processing method, storage medium and system for earth observation satellites, so as to solve the problem that most single solution algorithms in the prior art are prone to premature maturity and fall into local optimality. This scheme can improve the optimization performance of the algorithm and enhance the optimization performance and robustness of satellite observation mission planning problems.

[0007] The purpose of the present invention is achieved through the following technical solutions:

[0008] The first aspect of the present invention discloses a GA-SA hybrid optimization processing method for earth observation satellites. The hybrid optimization processing method is a two-layer parallel search structure for performing group parallel optimization: initializing the population, performing GA and SA searches in turn at various temperatures, wherein the initial solution of SA comes from the evolution result of GA, and the solution obtained by SA through the Metropolis sampling process is used as the initial population for the next round of GA evolution, and the cycle is iterated until convergence.

[0009] Preferably, the method comprises the following steps:

[0010] S1: define parameters, initialize the population and confirm the initial temperature;

[0011] S2: Evaluate the individuals in the current population and determine whether they have converged: if they have converged, output the optimization result; if they have not converged, proceed to step S3;

[0012] S3: GA is performed on individuals in the population for crossover, mutation and selection;

[0013] S4: Use the individuals generated in step S3 as the initial population of SA to perform SA search;

[0014] S5: Generate new individuals by the state generation function of SA;

[0015] S6: Perform constraint detection based on the constraint conditions during satellite observation and complete conflict resolution;

[0016] S7: accept new individuals with probability according to the Metropolis criterion;

[0017] S8: Perform constraint detection based on the constraint conditions during satellite observation and complete conflict resolution;

[0018] S9: Determine whether the SA sampling is stable: if not, return to step S5; if stable, return to step S2.

[0019] Preferably, the hybrid optimization processing method adopts binary coding, and the length of individuals in the population is the number of clustering tasks of the satellite in one orbital circle, then: 1 represents that the clustering task is executed in this circle, and 0 represents that it is not executed in this circle;

[0020] During the initialization process of the population, an individual with all chromosomes as 1 is generated, and then the individual is converted into a decimal system according to the gene length. The ratio of the converted value to the number of individuals is the coding interval between the two individuals. Other individuals in the population are generated incrementally based on the coding interval to obtain the initial population.

[0021] Preferably,

[0022] The crossover is as follows: randomly select two individuals from the population and calculate the fitness value, retain the individual with the larger fitness value, repeat the operation once to obtain two individuals with larger fitness values, and perform single-point crossover on these two individuals according to probability until all individuals in the population are selected;

[0023] The mutation described is: judging whether to mutate all individuals in the population according to probability, and randomly selecting mutation positions for the individuals to mutate to perform positional mutation;

[0024] The selection is as follows: the parent population and the new population generated by crossover and mutation are mixed, the individuals with the largest fitness in the two generations of populations are retained, two individuals are randomly selected from the remaining individuals and their fitness values ​​are calculated, the individuals with the larger fitness values ​​are retained, and the above operation is repeated until a new population is generated.

[0025] Preferably, the fitness value of an individual is the difference between the individual's objective function value and the lower bound of the fitness function; the objective function value is a weighted combination of the task priority function value and the satellite completion rate function value, the task priority function value is the ratio of the priority of tasks completed by the satellite to the priority of all tasks to be observed, and the satellite completion rate function value is the ratio of the number of tasks finally executed by the satellite to the number of all tasks to be observed.

[0026] Preferably, the constraints include: observation time window, observation task arrangement, preparation time, roll angle conversion time, single task observation duration, single maximum power-on time, maximum power-on times per lap and satellite energy.

[0027] Preferably, the SA determines whether the SA sampling is stable by determining whether the temperature drops to a set threshold, and the GA determines whether it converges by determining whether the difference between the maximum fitness value and the average fitness value of individuals in the population is less than a set threshold.

[0028] Preferably, the parameters include satellite orbit parameters, satellite detector parameters and algorithm parameters of GA and SA.

[0029] A second aspect of the present invention discloses a storage medium, wherein the storage medium stores a program, and when the program is executed by a processor, the GA-SA hybrid optimization processing method for earth observation satellites as described above is implemented.

[0030] The third aspect of the present invention discloses a GA-SA hybrid optimization processing system for earth observation satellites, at least one processor, and a memory communicatively connected to the at least one processor; the memory stores a program for implementing any of the GA-SA hybrid optimization processing methods for earth observation satellites as described above, and the at least one processor is configured to execute the program stored in the memory.

[0031] The working principle of the present invention is:

[0032] Compared with the prior art, the present invention has the following beneficial effects:

[0033] By constructing a hybrid algorithm of genetic algorithm and simulated annealing algorithm, the optimization performance of the algorithm can be improved as a whole, and the advantages of each single algorithm can be brought into play. At the same time, the two algorithms can make up for their respective shortcomings after being mixed, thereby improving the optimization performance and robustness of satellite observation mission planning problems. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 It is a schematic diagram of the process of GA-SA hybrid optimization processing method;

[0035] Figure 2 Schematic diagram of a single-point crossover in the crossover process of this method;

[0036] Figure 3 This is a schematic diagram of chromosome mutation during the mutation process in this method;

[0037] Figure 4 This is the fitness change curve when the number of tasks in the application example is set to 50;

[0038] Figure 5 This is the fitness change curve when the number of tasks in the application example is set to 100;

[0039] Figure 6 This is the fitness change curve when the number of tasks in the application example is set to 150;

[0040] Figure 7 This is a comparison chart of the fitness mean and variance of the three algorithms: GA, SA and GA-SA. DETAILED DESCRIPTION

[0041] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments.

[0042] Satellite mission planning is a complex optimization problem with multiple combinations and multiple objectives. The optimization results of traditional algorithms are often not ideal, and the improvement degree of a single algorithm has certain limitations. The present invention constructs a hybrid algorithm of genetic algorithm (GA) and simulated annealing algorithm (SA) to improve the optimization performance of the algorithm, which can give full play to the advantages of each single algorithm.

[0043] The GA-SA hybrid algorithm is a two-layer parallel search structure. The hybrid algorithm performs GA and SA searches in sequence at different temperatures. The initial solution of SA comes from the evolution result of GA, and the solution obtained by SA through the Metropolis sampling process becomes the initial population for further evolution of GA. At the spatial level, GA provides a parallel search structure, which transforms SA into a parallel SA algorithm. Therefore, the hybrid algorithm always performs group parallel optimization.

[0044] Example

[0045] This embodiment discloses a GA-SA hybrid optimization processing method for an earth observation satellite. Figure 1 As shown, the following steps are included:

[0046] 1. Given the algorithm parameters, initialize the population and confirm the initial temperature;

[0047] 2. Evaluate the current population individuals and determine whether they have converged;

[0048] 3. If convergence has been achieved, the optimization result is output; otherwise, the genetic algorithm is executed to select crossover mutation;

[0049] 4. The individuals generated by the genetic algorithm after selection, crossover and mutation are used as the initial solution of the simulated annealing algorithm to obtain the initial population of the simulated annealing algorithm and perform simulated annealing algorithm search;

[0050] 5. Generate new individuals by the state generation function of the simulated annealing algorithm. By introducing the constraint conditions during satellite observation, that is, the width of the clustering task time window must be greater than the maximum observation time requirement in the meta-task that forms this clustering task, constraint detection is performed to complete conflict elimination;

[0051] 6. Adopt the Metropolis criterion to accept new individuals with probability; perform constraint detection again through the constraints during satellite observation to complete conflict elimination;

[0052] 7. Determine whether the simulated annealing algorithm sampling is stable. If not, return to generate new individuals by the simulated annealing algorithm state generation function to cycle until it is stable. If stable, execute the cooling process and perform the next GA-SA algorithm iteration.

[0053] 8. Determine the termination condition and stop the loop iteration when the termination condition is reached.

[0054] The GA-SA hybrid algorithm coding rule uses binary coding. The length of each individual in the population is the number of clustering tasks of the satellite in one orbital cycle. The binary coded 1 represents that the clustering task is executed in this cycle, and 0 represents that it is not executed in this cycle. The process of generating the initial solution of the genetic algorithm generates an individual with all chromosomes 1 to ensure that there is an individual with the best fitness in the initial solution. Then the individual with all 1 is converted to the corresponding decimal according to the gene length. The ratio of this value to the number of individuals is the coding interval between the two individuals. The other individuals in the population are generated incrementally according to this value, so that the initial solution of the entire GA-SA hybrid algorithm is generated. The initial solution of the simulated annealing algorithm is the individual generated by the genetic algorithm after selection, crossover and mutation.

[0055] According to the above, the hybrid algorithm focuses on observing tasks with higher priority and completing as many tasks as possible. The optimization goal is to maximize the task priority and the number of completed tasks. Therefore, the optimization objective function Goal is selected as a combined optimization goal, which is a weighted combination of task priority and the number of observed tasks. The calculation formula is:

[0056] maxGoal=a·PFR+b·NFR

[0057] Among them, the priority function PFR, the function value is equal to the ratio of the priority of the tasks completed by the satellite to the priority of all the tasks to be observed; the satellite completion rate function is NFR, and its function value is the ratio of the number of tasks finally executed by the satellite to the number of all the tasks to be observed; a and b are weighted coefficients respectively, satisfying a+b=1.

[0058] Specifically, the genetic operations in the GA-SA hybrid algorithm include:

[0059] 1. Fitness function selection:

[0060] The fitness of an individual refers to the measure of the degree of advantage of the individual in the survival of the population, which is used to distinguish the "good" and "bad" individuals. The fitness is calculated using the fitness function. The fitness function determines the fitness of an individual through individual characteristics. There is a relationship between the fitness function and the objective function. For the optimization goal selected by this hybrid algorithm, "maximization" is the goal. The fitness function can be designed as follows:

[0061] f=Goal-f min

[0062] Among them, f is the individual fitness, Goal is the objective function value of the individual, and f min is the lower bound of the fitness function.

[0063] 2. Crossover operation:

[0064] Randomly select two individuals from the population, calculate the fitness values ​​of the two individuals, retain the individual with the larger fitness value, and perform the same operation again to obtain two individuals with higher fitness values. These two individuals are crossovered according to probability until all individuals are selected. The probability of crossover p c As shown in the following formula:

[0065]

[0066] In the above formula, p c1 =0.9, p c2 =0.6, f b is the larger fitness value of the two individuals, f max is the maximum fitness value in the population, f avg is the average fitness of the population. This adaptive crossover probability can prevent the algorithm from premature early maturity and is conducive to the later evolution of the algorithm. Single-point crossover is used during crossover, such as Figure 2 shown.

[0067] 3. Mutation operation:

[0068] Mutation is to change one or some gene values ​​to other allele values ​​for each individual in the population with a certain probability. Mutation can enhance the local search ability of the algorithm. According to different encoding methods, there are real-value mutation and binary mutation. Binary mutation only needs to negate the corresponding gene value, while real-value mutation replaces the corresponding gene value with other random values ​​within the value range. The general steps of the mutation operation are: first determine whether to mutate all individuals in the population according to the pre-set mutation probability, and then randomly select the mutation bit for the mutated individuals to mutate.

[0069] Individual mutation probability p m A similar calculation formula as the crossover probability is also used:

[0070]

[0071] In the above formula, p m1 =0.1, p m2 =0.001, f c is the fitness of the individual to be mutated, f max is the maximum fitness value in the population, f avg is the average fitness in the population. The selected mutation method is positional mutation, such as Figure 3 shown.

[0072] 4. Select an operation:

[0073] Using certain rules or methods, some excellent individuals are selected from the current population according to the fitness of each individual and inherited into the next generation population. The parental population is mixed with the new population generated by crossover and mutation for selection. The specific operation is as follows: Among the two generations of populations, the individual with the highest fitness is directly retained. For the other individuals, two individuals are randomly selected, the fitness values of the two are calculated, and the individual with the larger fitness value is retained. This operation is repeated until a new population is generated. By calculating the fitness of each individual in the population, the quality of the individual is judged. The proportion of the task priority completed by the satellite in all task priorities and the proportion of the number of tasks completed by the satellite during observation in all task numbers are set as the optimization objectives for fitness value calculation.

[0074] Specifically, the generation of a new solution by the simulated annealing algorithm SA is as follows:

[0075] For an individual at a certain temperature, a new solution is generated by a new solution generation function, and the acceptance criterion for the new solution adopts the Metropolis criterion:

[0076]

[0077] E2 is the energy of the newly generated solution, and E1 is the energy of the original solution, which is defined as the negative of the fitness value; if E2 < E1, the newly generated solution is accepted, otherwise with a probability the newly generated solution is accepted, where T is the temperature; when the new solution is determined to be accepted, the new solution replaces the current solution, and at the same time the objective function value is corrected. The current individual realizes one iteration at a certain temperature, and continues the next iteration until the number of iterations is satisfied, and then the same operation is performed on the next individual. After all individuals in the population have completed the iteration at this temperature, the temperature reduction process is executed for the next GA - SA hybrid algorithm iteration.

[0078] The solution generated by the GA - SA hybrid algorithm may not satisfy the constraint conditions during satellite observation. It is necessary to perform a constraint check on the generated solution, comprehensively considering the satellite visible time window, no crossover of detector observation tasks, preparation time for the detector to observe different tasks, time for slewing angle conversion, observation duration of the task itself, maximum single - turn power - on time, maximum number of power - on times per orbit, and satellite energy constraint, to judge whether there are conflicts when these tasks are executed. If any of the constraint conditions are not met, the current solution needs to be corrected to eliminate the conflicting tasks. In the designed algorithm, two conflict resolution checks are performed to improve the stability of the algorithm. The following gives the mathematical expressions of the above - mentioned constraints:

[0079] 1) Observation time window constraint:

[0080]

[0081] Among them, represents the start time of observation task u, Indicates the task end time, d u Indicates the observation time, Represents the set of all feasible time windows for a task.

[0082] 2) A probe cannot perform two tasks at the same time, and the satellite cannot overlap when performing any two tasks:

[0083]

[0084] x is a decision variable, representing whether to execute task u after task v is completed. If so, then x uv =1, otherwise, x uv =0.

[0085] 3) The detector has a certain amount of preparation time for performing two observation missions, mainly considering the maneuvering time of the satellite sensor and the preparation time of the imaging sensor:

[0086]

[0087] Among them, s uv The preparation time from task u to task v, TransT uv is the roll angle conversion time.

[0088] 4) Side swing angle conversion time constraint:

[0089]

[0090] Among them, Δθ uv It represents the change of the observed roll angle of the two tasks, and v is the average roll angular velocity.

[0091] 5) Satellite observation duration constraints:

[0092]

[0093] 6) Satellite detectors have a maximum power-on time constraint:

[0094] maxT≥d u ,u={1,2,…,N}

[0095] 7) Maximum boot times constraint:

[0096]

[0097] 8) Satellite energy constraints

[0098]

[0099] ei represents the energy consumed per unit time during observation, ε uv It represents the energy consumed per unit time for the probe to adjust and convert from mission u to mission v, and E is the total energy of the satellite.

[0100] There are two termination conditions for the GA-SA hybrid algorithm: one is that the temperature of the simulated annealing algorithm drops to the required temperature eps, then the cycle stops; the other stopping condition is that the difference between the maximum and average individual fitness in the population at a certain temperature is less than a certain value eps, then it stops.

[0101] Application Examples

[0102] Satellite orbit parameters:

[0103]

[0104] Satellite detector parameters:

[0105] Field of view FoV(°) Maximum working time maxT(s) Maximum side swing angle maxθ(°) 10 150 ±45

[0106] Given the constraint parameters:

[0107] ei <![CDATA[ε uv ]]> E CountMax 1.0 0.5 1200 60

[0108] Given the algorithm parameters of GA-SA:

[0109]

[0110] The fitness change curves of the GA-SA hybrid algorithm when the number of original tasks is 50, 100, and 150 are given respectively, as shown in Figure 4 , Figure 5 , Figure 6 As shown in Figure 2, the demand side swing angle, observation time, and priority of each task are randomly given. According to the simulation result diagram, the fitness value of the first iteration is higher. This is because there are optimal individuals without constraint checks during the initial population generation process, so the calculated fitness value is higher.

[0111] The optimal fitness values ​​achieved in the end are different when the number of tasks is different, and the more original tasks there are, the lower the fitness value achieved in the end. When the initial number of tasks is 50, the final fitness value is 0.6294. When the initial number of tasks is 100, the final fitness value is 0.3718. When the initial number of tasks is 150, the final fitness value is 0.2833.

[0112] Figure 7When the number of original tasks is different, the GA-SA hybrid algorithm is compared with the GA algorithm and the SA algorithm. Each algorithm is run ten times when the number of original tasks is 50, 100, and 150, and the average and variance of the fitness values ​​of the three algorithms are compared. When the number of original tasks is 50, the average fitness value is about 0.63202, and the variance value is about 3.262×10 -7 When the initial number of tasks is 100, the average value of fitness is about 0.3712 and the variance is about 1.343×10 -7 When the number of original tasks is 150, the average fitness value is about 0.28974 and the variance value is about 3.846×10 -7 From the simulation results, it can be seen that the GA-SA hybrid algorithm can obtain better solutions than the other two algorithms, and the average fitness value is higher. The variance of the GA-SA algorithm is lower than that of the other two algorithms, and the algorithm stability is stronger.

[0113] The above description of the embodiments is to facilitate the understanding and use of the invention by those skilled in the art. It is obvious that those skilled in the art can easily make various modifications to these embodiments and apply the general principles described herein to other embodiments without creative work. Therefore, the present invention is not limited to the above embodiments, and improvements and modifications made by those skilled in the art based on the disclosure of the present invention without departing from the scope of the present invention should be within the scope of protection of the present invention.

Claims

1. A GA-SA hybrid optimization processing method for an earth observation satellite, characterized in that: The hybrid optimization processing method is a two-layer parallel search structure for group parallel optimization: the population is initialized, and GA and SA searches are performed in turn at different temperatures, wherein the initial solution of SA comes from the evolution result of GA, and the solution obtained by SA through the Metropolis sampling process is used as the initial population for the next round of GA evolution, and the cycle is iterated until convergence.

2. The GA-SA hybrid optimization processing method for an earth observation satellite according to claim 1, characterized in that: The steps include: S1: define parameters, initialize the population and confirm the initial temperature; S2: Evaluate the individuals in the current population and determine whether they have converged: if they have converged, output the optimization results; If it has not converged, proceed to step S3; S3: GA is performed on individuals in the population for crossover, mutation and selection; S4: Use the individuals generated in step S3 as the initial population of SA to perform SA search; S5: Generate new individuals by the state generation function of SA; S6: Perform constraint detection based on the constraint conditions during satellite observation and complete conflict resolution; S7: accept new individuals with probability according to the Metropolis criterion; S8: Perform constraint detection based on the constraint conditions during satellite observation and complete conflict resolution; S9: Determine whether the SA sampling is stable: if not, return to step S5; if stable, return to step S2.

3. The GA-SA hybrid optimization processing method for an earth observation satellite according to claim 2, characterized in that: The hybrid optimization processing method adopts binary coding, and the length of individuals in the population is the number of clustering tasks of the satellite in one orbital circle, then: 1 represents that the clustering task is executed in this circle, and 0 represents that it is not executed in this circle; During the initialization process of the population, an individual with all chromosomes as 1 is generated, and then the individual is converted into a decimal system according to the gene length. The ratio of the converted value to the number of individuals is the coding interval between the two individuals. Other individuals in the population are generated incrementally based on the coding interval to obtain the initial population.

4. The GA-SA hybrid optimization processing method for an earth observation satellite according to claim 2, characterized in that: The crossover is as follows: randomly select two individuals from the population and calculate the fitness value, retain the individual with the larger fitness value, repeat the operation once to obtain two individuals with larger fitness values, and perform single-point crossover on these two individuals according to probability until all individuals in the population are selected; The mutation described is: judging whether to mutate all individuals in the population according to probability, and randomly selecting mutation positions for the individuals to mutate to perform positional mutation; The selection is as follows: the parent population and the new population generated by crossover and mutation are mixed, the individuals with the largest fitness in the two generations of populations are retained, two individuals are randomly selected from the remaining individuals and their fitness values ​​are calculated, the individuals with the larger fitness values ​​are retained, and the above operation is repeated until a new population is generated.

5. The GA-SA hybrid optimization processing method for an earth observation satellite according to claim 4, characterized in that: The fitness value of an individual is the difference between the individual's objective function value and the lower bound of the fitness function; The objective function value is a weighted combination of the task priority function value and the satellite completion rate function value, wherein the task priority function value is the ratio of the priority of tasks completed by the satellite to the priority of all tasks to be observed, and the satellite completion rate function value is the ratio of the number of tasks finally executed by the satellite to the number of all tasks to be observed.

6. The GA-SA hybrid optimization processing method for an earth observation satellite according to claim 2, characterized in that: The constraints include: observation time window, observation task arrangement, preparation time, roll angle conversion time, single task observation time, single maximum startup time, single maximum startup times and satellite energy.

7. The GA-SA hybrid optimization processing method for an earth observation satellite according to claim 2, characterized in that: The SA determines whether the SA sampling is stable by determining whether the temperature drops to a set threshold, and the GA determines whether it converges by determining whether the difference between the maximum fitness value and the average fitness value of individuals in the population is less than a set threshold.

8. The GA-SA hybrid optimization processing method for an earth observation satellite according to claim 2, characterized in that: The parameters include satellite orbit parameters, satellite detector parameters and GA and SA algorithm parameters.

9. A storage medium, characterized in that: The storage medium stores a program, and when the program is executed by a processor, the GA-SA hybrid optimization processing method for an earth observation satellite as described in any one of claims 1-8 is implemented.

10. A GA-SA hybrid optimization processing system for earth observation satellites, characterized in that: At least one processor and a memory communicatively connected to the at least one processor; the memory stores a program for implementing the GA-SA hybrid optimization processing method for earth observation satellites as described in any one of claims 1-8, and the at least one processor is configured to execute the program stored in the memory.

Images