Method, device and storage medium for planning multi-agile satellite observation of dense target clusters

Abstract

The present application relates to a kind of multi-agile satellite observation dense group target planning method, equipment and storage medium, obtain all agile satellite parameter information and observation group target information, construct multi-agile satellite task planning evaluation function and constraint condition;The visibility of agile satellite to observation target is calculated, and the visible task sequence is constructed according to the visibility, and the visible task sequence is encoded based on cuckoo algorithm;Using cuckoo algorithm, the observation scheme of multi-agile satellite observation dense group target is output.The present application can greatly improve the convergence speed of task planning and effectively inhibit algorithm precocity, to realize the rapid and efficient planning of multi-agile satellite and dense group target.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of satellite observation, and in particular to a multi-agile satellite observation dense cluster target planning method, device and storage medium. BACKGROUND

[0002] Compared with traditional non-agile observation satellites, agile earth observation satellites can realize observation of a target when the satellite has not reached the target point or has flown past the target point by adjusting the satellite attitude, greatly extending the satellite's visible window to the target, and enabling the agile satellite to have stronger earth observation capability.

[0003] Due to the added pitch and yaw degrees of freedom of the agile satellite, the flexible observation mode brought by the multi-degree-of-freedom and high maneuverability of the agile satellite makes the solution space of the agile imaging satellite scheduling problem larger, and the planning and scheduling problem of a single agile satellite is already a time-dependent combinatorial optimization problem with high complexity. The collaborative planning and scheduling of multiple agile satellites involves both multi-satellite task allocation and collaborative planning and scheduling, and the coupling between the two greatly increases the solution space, especially for the observation of dense cluster targets by multiple agile satellites, which is a complex observation task with a large number of observations and strong observation coupling, and the problem is even more complex.

[0004] Currently, there is relatively little research on the collaborative scheduling of multiple agile satellites. The existing multi-satellite collaborative planning and scheduling problems mainly have three strategies; one is random allocation, which randomly allocates tasks to multiple satellites, which makes it difficult to ensure the excellence of the scheduling strategy; two is according to the time window, the earliest allocation, which allocates tasks to the satellite with the earliest visible time window, Bianchessi allocates tasks to the satellite with the earliest visible time window, which is applied to the planning and scheduling of the COSMO-SkyMed constellation; Richards et al. allocate observation tasks to each satellite according to the order of the visible window, so although it is multi-satellite scheduling, the satellites are independent of each other and do not consider the observation conflicts between multiple satellites; three is not to allocate tasks, but to treat the multi-satellite problem as a single-satellite multi-orbit problem for overall optimization, which is only suitable for small-scale number of satellites and cannot solve the planning and scheduling of large-scale number of agile satellites. At the same time, the general collaborative optimization algorithms currently used, such as constraint programming, greedy, dynamic programming, tabu search algorithm, improved genetic algorithm, ant colony algorithm, genetic simulated annealing hybrid algorithm, etc. However, when it comes to the planning and scheduling of multiple agile satellites observing dense cluster targets, which is a super-large solution space problem, it is easy to cause the algorithm to fall into a local optimal solution prematurely, and it is difficult to guarantee the optimization efficiency. SUMMARY

[0005] In view of the above technical problems, the present application provides a multi-agile satellite observation dense group target planning method, device and storage medium, which can greatly improve the convergence speed and effectively suppress the premature algorithm, so as to realize the rapid and efficient planning of multi-agile satellites and dense group targets.

[0006] The technical solution for achieving the object of the present application is as follows: a multi-agile satellite observation dense group target planning method, comprising the following steps:

[0007] Step S1, acquiring all agile satellite parameter information and observation group target information, constructing a multi-agile satellite task planning evaluation function and constraint conditions;

[0008] Step S2, calculating the visibility of agile satellites to observation targets, constructing a visible task sequence according to the visibility, and encoding the visible task sequence based on a cuckoo algorithm;

[0009] Step S3, using the cuckoo algorithm to output an observation scheme of multi-agile satellites observing dense group targets.

[0010] According to an aspect of the present application, in the step S1, Ns agile satellites and Nt observation targets are included, and the planning evaluation function is:

[0011]

[0012] Wherein, Ti represents the time consumed for observing the i-th observation target.

[0013] According to an aspect of the present application, in the step S1, the constraint conditions at least include agile satellite attitude maneuvering constraints, visible window execution constraints, agile satellite imaging observation time constraints, solar elevation angle constraints, and task condition constraints.

[0014] According to an aspect of the present application, in the step S2, the visible task sequence is encoded based on the cuckoo algorithm, and is specifically represented as:

[0015]

[0016] Wherein, represents a set of observation schemes for N observation targets obtained by the i-th bird nest in the t-th generation, Toc N represents the observation time of the Nth target obtained in the planning evaluation function.

[0017] According to an aspect of the present application, in the step S3, it specifically comprises:

[0018] Step S301, updating the cuckoo nest position by adjusting the step length and discovery probability of Lévy flight to optimize the task sequence;

[0019] Step S302, dynamically adjust the bird nest by learning strategy, discard the poor bird nest and generate a new bird nest;

[0020] Step S303, judge whether the termination condition is met, output the best bird nest;

[0021] Wherein, any egg in the bird nest of the cuckoo represents a task scheduling scheme, and any cuckoo has and only has one egg.

[0022] According to one aspect of the application, in the step S301, when the i-th cuckoo generates a new solution, the execution of Lévy flight is as follows:

[0023]

[0024] In the formula, α>0 represents a step control quantity, Num represents the number of hosts, t represents the current generation number, represents the i-th bird nest of the t-th generation, represents point-to-point multiplication, and Lévy(λ) represents a Lévy flight path, wherein Lévy~g -λ (1<λ≤3).

[0025] λ is a Lévy flight parameter, which is the average or expected value of the occurrence of events during a unit interval, and the calculation formula of Lévy is as follows:

[0026]

[0027] Wherein, μ and v are normal distribution parameters, θ μ and θ v are standard deviations of normal distribution,

[0028]

[0029] Wherein, θ v =1, and Γ represents a gamma function. Lévy flight provides random walk, and a random step is obtained from a Lévy distribution function.

[0030] According to one aspect of the application, in the step S302, the bird nest moves to the global optimal position, and the global optimal value (solution) in the current iteration is always greater than or at least equal to the global optimal value in the last iteration, F(x g (t)) and F(x g (t-1)) represent the optimal solution of the bird nest found in the current iteration and the last iteration respectively, including:

[0031] The evolution degree factor of the learning cuckoo nest is defined, and the formula is as follows: The bird nest evolution degree factor utilizes historical information of the algorithm and reflects the search speed of levy flight

[0032] wherein 0 d ≤1, as md decreases, the bird nest current best solution will approach the best solution;

[0033] The bird nest aggregation degree factor of the learning type cuckoo is defined, and the formula is:

[0034] wherein F(xg(t)) represents the optimal solution, and Mt represents the average value of the optimal positions of all bird nests, and the formula is: F(xi(t)) represents that in the current iteration, the optimal solution F(xg(t)) is superior to the function value of each bird nest.

[0035] According to one aspect of the application, the step length control quantity and the discovery probability are updated by the following formula:

[0036] α = α0- α m (m d ,a am )+ α s (s d ,a as )

[0037] p a = p0- p m (m d ,a pm )+ p s (s d ,a ps )

[0038] wherein α0 is an initial step length and p0 is an initial bird egg discovery probability,

[0039] By learning historical information and making the following response actions:

[0040]

[0041]

[0042]

[0043]

[0044] wherein l α is a step length efficacy factor, l p is a discovery probability efficacy factor, α m (m d ,a am ) and α s (s d ,aas ) respectively represent the effect of the bird nest evolution degree factor and the bird nest aggregation degree factor on the step size a, p m (m d ,a pm ) and p s (s d ,a ps ) respectively represent the effect of the bird nest evolution degree factor and the bird nest aggregation degree factor on the discovery probability p a .

[0045] According to an aspect of the present application, there is provided an apparatus comprising one or more processors, one or more memories, and one or more computer programs; wherein the processor is connected with the memory, the one or more computer programs are stored in the memory, and when the apparatus is running, the processor executes the one or more computer programs stored in the memory, so that the apparatus executes a planning method for observing a dense group target by a plurality of agile satellites as described in any one of the above technical solutions.

[0046] According to an aspect of the present application, there is provided a computer readable storage medium for storing computer instructions, which, when executed by a processor, implement a planning method for observing a dense group target by a plurality of agile satellites as described in any one of the above technical solutions.

[0047] According to the concept of the present application, a planning method for observing a dense group target by a plurality of agile satellites, an apparatus, and a computer readable storage medium are proposed, a learning-type cuckoo search algorithm is first proposed to solve the super-large solution space problem of planning a plurality of agile satellites and a dense group target, a planning and scheduling model for observing a dense group target by a plurality of agile satellites is established, each agile satellite in the plurality of agile satellites and each observation target can be cooperatively optimized, the step size and the bird nest discovery probability of the cuckoo search algorithm are adaptively adjusted by learning the relationship between the cuckoo search parameters and the calculation convergence speed, the task planning convergence speed can be greatly improved and the algorithm prematureness can be effectively inhibited by the present application, so as to realize rapid and efficient planning of a plurality of agile satellites and a dense group target. BRIEF DESCRIPTION OF DRAWINGS

[0048] Figure 1 a flowchart schematically showing a planning method for observing a dense group target by a plurality of agile satellites according to an embodiment of the present application;

[0049] Figure 2 a flowchart schematically showing a planning method for observing a dense group target by a plurality of agile satellites according to an embodiment of the present application;

[0050] Figure 3 a distribution diagram schematically showing 100 dense group targets according to an embodiment of the present application;

[0051] Figure 4 Fig. 1 schematically shows a distribution of 100 dense cluster targets according to an embodiment of the present application;

[0052] Figure 5 Fig. 5 schematically shows a planning and scheduling result of 50 dense cluster targets after using the present application according to an embodiment of the present application;

[0053] Figure 6 Fig. 6 schematically shows a planning and scheduling result of 100 dense cluster targets after using the present application according to an embodiment of the present application. DETAILED DESCRIPTION

[0054] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments. Obviously, the drawings in the following description only represent some embodiments of the present application, and other drawings can also be obtained by those skilled in the art without any creative effort based on these drawings.

[0055] The present application will be described in detail below in combination with the drawings and specific embodiments. The embodiments cannot be described one by one here, but the embodiments of the present application are not limited to the following embodiments.

[0056] As shown in Fig. 1, a planning method for observing dense cluster targets by multiple agile satellites according to an embodiment of the present application comprises the following steps: Figures 1 to 6

[0057] Step S1, acquiring all the parameter information of agile satellites and the information of observed cluster targets, and constructing a planning and scheduling evaluation function and constraint conditions for the multiple agile satellites;

[0058] Step S2, calculating the visibility of the agile satellites to the observed targets, constructing a visible task sequence according to the visibility, and encoding the visible task sequence based on a cuckoo algorithm;

[0059] Step S3, outputting an observation scheme for observing dense cluster targets by multiple agile satellites by using the cuckoo algorithm.

[0060] In this embodiment, a planning and scheduling model for observing dense cluster targets by a multiple agile satellite system is established, each agile satellite in the multiple agile satellite system and each observed target can be cooperatively optimized, the step length and discovery probability of Lévy flight are adaptively adjusted by learning the current state and historical state of the bird nest, the depth and breadth of the search can be balanced to accelerate the convergence speed and jump out of the local optimal solution, so that the global optimum can be quickly searched in a larger solution space, the convergence speed of task planning can be greatly improved, and the algorithm precocity can be effectively suppressed, so as to realize fast and efficient planning of multiple agile satellites and dense cluster targets. ​

[0061] In one embodiment of the present application, preferably, in the step S1, Ns agile satellites, Nt observation targets are included, and the planning evaluation function is:

[0062]

[0063] wherein Ti represents the time spent on observing the ith observation target.

[0064] In one embodiment of the present application, preferably, in the step S1, the constraint conditions include at least agile satellite attitude maneuver constraint, visible window execution constraint, agile satellite imaging observation time constraint, sun elevation angle constraint, and task condition constraint.

[0065] wherein the agile satellite attitude maneuver constraint is expressed as, if the same satellite observes adjacent 2 targets, the actual observation start time between the two observation targets i, j is greater than or equal to the sum of the maneuver time and the imaging time between the two targets, which is expressed as:

[0066]

[0067] The attitude maneuver time required by the agile satellite for observing the target i is expressed as , in seconds, is the time required for observing the target i.

[0068] The visible window execution constraint is expressed as, the satellite can only work in the visible execution window with the observation target point, is the observation start time of the target i with the satellite j in the kth orbit circle, is the observation end time of the target i with the satellite j in the kth orbit circle, and is the visible start time of the target i and the satellite j, is the visible end time of the target i and the satellite j, and the expression is:

[0069]

[0070] The agile satellite payload single-orbit longest working time constraint is expressed as:

[0071]

[0072] j represents the jth satellite; Cj is the number of orbit circles of the satellite j in the planning period, Kj is the maximum length of the satellite j that can work in a single orbit circle, and is the observation time of the target i; the agile satellite imaging observation time constraint is expressed as: ​​

[0073]

[0074] The solar elevation angle constraint is that the imaging visible light camera should be greater than the minimum solar elevation angle required by the ground observation when imaging the ground, and the expression is:

[0075] η Qi ≥η min ,

[0076] Wherein, η min is the minimum solar elevation angle.

[0077] The task condition constraint is that only one satellite can perform one task at one time to ensure observation quality, and the expression of the task condition constraint is:

[0078]

[0079] In an embodiment of the present application, preferably, in the step S2, the visible task sequence is encoded based on the cuckoo algorithm, and the specific expression is:

[0080]

[0081] In the method for observing dense cluster targets of multiple agile satellites based on cuckoo, each egg in the bird nest of the cuckoo represents a task scheduling scheme, and each cuckoo can only produce one egg (thus representing a solution), The expression indicates a set of observation schemes about N observation targets obtained by the ith bird nest in the t generation, Toc N The expression indicates the observation time of the Nth target obtained in the planning evaluation function.

[0082] In an embodiment of the present application, preferably, in the step S3, the following steps are specifically included:

[0083] Step S301, updating the cuckoo nest position by adjusting the step length and discovery probability of Lévy flight to optimize the task sequence;

[0084] Step S302, dynamically adjusting the bird nest by learning strategy, discarding the poor bird nest and generating a new bird nest;

[0085] Step S303, judging whether the termination condition is met, and outputting the best bird nest;

[0086] Wherein, any egg in the bird nest of the cuckoo represents a task scheduling scheme, and any cuckoo has and only has one egg.

[0087] In this embodiment, the basic idea of the cuckoo optimization algorithm is based on the cuckoo's nest parasitic behavior and the bird's Lévy flight behavior, which contains three elements: selecting the optimal, taking local random movement, and selecting randomly through global Lévy flight. Lévy flight is a typical non-Gaussian random walk mechanism, and its flight second moment diverges, so that Lévy motion process often occurs in a small aggregation case. The change of flight path follows the heavy-tailed distribution, which can effectively avoid falling into the local optimal solution. At the same time, since the number of available host nests is fixed, the host discovers foreign eggs with a discovery probability of pa[0, 1] to simulate the real cuckoo bionics process, and the global optimization rhythm is adjusted at the same time.

[0088] In one embodiment of the present application, preferably, in the step S301, when the i-th cuckoo generates a new solution, the Lévy flight is performed as follows:

[0089]

[0090] In the formula, α>0 represents a step control quantity, Num represents the number of hosts, t represents the current generation number, represents the i-th nest of the t-th generation, represents point-to-point multiplication, and Lévy(λ) represents the Lévy flight path, wherein Lévy~g -λ (1<λ≤3).

[0091] λ is a Lévy flight parameter, which is the average or expected value of the event occurrence in a unit interval. The calculation formula of Lévy is as follows:

[0092]

[0093] wherein μ and v are normal distribution parameters, θ μ and θ v are the standard deviations of the normal distribution,

[0094]

[0095] wherein, θ v =1, and Γ represents the gamma function. Lévy flight provides random walk, and the random step is obtained from the Lévy distribution function.

[0096] In one embodiment of the present application, preferably, in the step S302, in the learning cuckoo algorithm, the nest moves to the global optimal position, and the global optimal value (solution) in the current iteration is always greater than or at least equal to the global optimal value in the last iteration, F(x g (t)) and F(x gF(xg(t)) and F(xg(t-1)) represent the optimal solution of the nest found in the current iteration and the last iteration, respectively, including:

[0097] The nest evolution factor of the learning cuckoo is defined as: The nest evolution factor uses the historical information of the algorithm and reflects the search speed of levy flight,

[0098] Wherein, 0 d ≤1, as md decreases, the current optimal solution of the nest will approach the optimal solution;

[0099] The nest aggregation factor of the learning cuckoo is defined as: The aggregation factor represents the aggregation degree of all current nests, which embodies the diversity of the nests, and the greater the aggregation factor, the smaller the diversity of the nests.

[0100] Wherein, F(xg(t)) represents the optimal solution, and Mt represents the average value of the optimal positions of all nests, and the formula is: F(xg(t)) represents that in the current iteration, the optimal solution F(xg(t)) is better than the function value of each nest.

[0101] In an embodiment of the present application, preferably, the step control quantity and the discovery probability are updated by the following formula:

[0102] α = α0- α m (m d ,a am )+ α s (s d ,a as )

[0103] p a = p0- p m (m d ,a pm )+ p s (s d ,a ps )

[0104] Wherein, α0 is the initial step size and p0 is the initial probability of finding an egg,

[0105] By learning the historical information and making the following response actions:

[0106]

[0107]

[0108]

[0109]

[0110] wherein, l α is a step length efficacy factor, l p is a discovery probability efficacy factor, a m (m d , a am ) and a s (s d , a as ) represent the influence effect of the bird nest evolution degree factor and the bird nest aggregation degree factor on the step length a, respectively, p m (m d , a pm ) and p s (s d , a ps ) represent the influence effect of the bird nest evolution degree factor and the bird nest aggregation degree factor on the discovery probability p a , respectively.

[0111] In order to increase the diversity of the solution, the learning cuckoo algorithm weakens the influence effect of the evolution degree factor at the initial stage of the algorithm, and strengthens the influence effect of the aggregation degree factor, in order to improve the convergence speed of the optimization process, the influence of the aggregation degree factor response should be gradually increased, and the influence of the evolution degree factor action should be gradually reduced.

[0112] In an embodiment of the present application, preferably, in step S303, the termination condition can be that the set maximum number of iterations is reached or the iteration error is less than e dd (e dd may be set according to the task requirement), when the termination condition is not met, return to step S301.

[0113] As shown in FIG. 1, a multi-agile satellite observation dense group target planning method according to the present application comprises the following steps: Figure 2 After obtaining the multi-agile satellite parameter information and the observation group target information, the multi-agile satellite task planning evaluation function and the constraint condition are established, the visibility of the agile satellite to the target is calculated, the visible task sequence is encoded, the learning cuckoo algorithm parameters are initialized, the bird nest is randomly obtained, the bird nest position is updated by Lévy flight to optimize the task sequence, the bird nest is dynamically adjusted by the learning strategy, the poor bird nest is discarded and a new bird nest is generated, and it is judged whether the termination condition is met, if yes, the best bird nest is output; if no, return to the step of "randomly obtaining the bird nest, updating the bird nest position by Lévy flight to optimize the task sequence".

[0114] The specific implementation is as follows:

[0115] In the following table, Table 1 is the satellite parameters of an observation system composed of 12 agile optical satellites. The maneuvering capability of each satellite is 60°.

[0116]

[0117] Table 1

[0118] Figure 5 and Figure 6 are simulation comparison results of 50 dense cluster targets and 100 dense cluster targets by using the planning algorithm of the application and the classic improved cuckoo search algorithm (ICS), quantum particle swarm optimization algorithm (QPSO) and genetic algorithm (GA) respectively.

[0119] From the simulation results, it can be seen that:

[0120] (1) The application has a faster convergence speed. For 50 dense cluster targets, it can converge at 896 iterations, reach a better convergence position, and other methods need more than 2500 iterations; for 100 dense cluster targets, the application can converge at 2374 iterations and reach a better convergence position, and other methods need more than 2800 iterations.

[0121] (2) The application can obtain a better planning scheme. For 50 dense cluster targets, the planning and scheduling scheme using the application can observe all targets in 32.61 hours, and the corresponding ICS, QPSO and GA methods need 36.0 hours, 46.96 hours and 51.64 hours respectively. For 100 dense cluster targets, the planning and scheduling scheme using the application can observe all targets in 69.57 hours, and the corresponding ICS, QPSO and GA methods need 77.70 hours, 81.54 hours and 104.469 hours respectively.

[0122] According to an aspect of the application, there is provided a device comprising one or more processors, one or more memories, and one or more computer programs; wherein the processor is connected with the memory, and the one or more computer programs are stored in the memory, and when the device is running, the processor executes the one or more computer programs stored in the memory, so that the device executes a planning method for observing dense cluster targets by multiple agile satellites according to any one of the above technical solutions.

[0123] According to an aspect of the application, there is provided a computer readable storage medium for storing computer instructions, which are executed by a processor to implement a planning method for observing dense cluster targets by multiple agile satellites according to any one of the above technical solutions.

[0124] In summary, the present application proposes a multi-agile satellite observation dense group target planning method, device and computer readable storage medium, firstly proposes a learning type cuckoo search algorithm to solve the super large solution space problem of multi-agile satellite and dense group target planning, establishes a multi-agile satellite system observation dense group target planning and scheduling model, can cooperatively optimize each agile satellite in the multi-agile satellite system and each observation target, through learning the relationship between cuckoo parameters and calculation convergence speed, adaptively adjusts the step length and bird nest discovery probability of the cuckoo algorithm, through the present application, the task planning convergence speed can be greatly improved and the algorithm precocity can be effectively inhibited, so that the multi-agile satellite and dense group target can be quickly and efficiently planned.

[0125] In addition, it should be noted that the present application can be provided as a method, apparatus or computer program product. Therefore, the embodiments of the present application can adopt a completely hardware embodiment, a completely software embodiment or an embodiment combining software and hardware aspects. Moreover, the embodiments of the present application can adopt the form of a computer program product implemented on one or more computer usable storage media containing computer usable program codes.

[0126] Embodiments of the present application are described with reference to flowcharts and / or block diagrams according to the method, terminal device (system) and computer program product of the embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and the combination of the flows and / or blocks in the flowcharts and / or block diagrams can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, embedded processor or other programmable data processing terminal device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing terminal device produce a means for implementing the functions specified in the flowcharts and / or block diagrams. Figure 1 one or more flows and / or blocks Figure 1 an apparatus for performing the functions specified in one or more flows and / or blocks.

[0127] These computer program instructions can also be stored in a computer readable memory capable of guiding the computer or other programmable data processing terminal device to work in a specific way, so that the instructions stored in the computer readable memory produce a product including instruction apparatus, which implements the functions specified in the flowcharts and / or block diagrams. Figure 1 one or more flows and / or blocks Figure 1 an apparatus for performing the functions specified in one or more flows and / or blocks. These computer program instructions can also be loaded into a computer or other programmable data processing terminal device, so that a series of operation steps are performed on the computer or other programmable terminal device to produce a computer implemented process, so that the instructions executed on the computer or other programmable terminal device provide a means for implementing the functions specified in the flowcharts and / or block diagrams. Figure 1 one or more flows and / or blocks Figure 1steps of a function specified in one or more blocks.

[0128] It is also to be noted that, as used herein, the terms "comprise", "comprising", or any other variation thereof are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can also include other elements not expressly listed or inherent to such process, method, article, or apparatus. An element proceeded by "comprises... a" does not, without more constraints, exclude the presence of additional identical elements in the process, method, article, or apparatus that comprises the element.

[0129] Finally, it is to be noted that the above-mentioned embodiments illustrate rather than limit the application, since the skilled person, having the benefit of the present disclosure will be able to devise suitable modifications and equivalents without departing from the scope of the application. Therefore, the appended claims are intended to cover all such modifications and equivalents.

Claims

1. A planning method for densely clustered targets observed by multiple agile satellites, comprising the following steps: Step S1: Obtain all agile satellite parameter information and observation group target information, and construct multi-agile satellite mission planning evaluation functions and constraints; Step S2: Calculate the visibility of the agile satellite to the observed target, construct a visible task sequence based on the visibility, and encode the visible task sequence based on the Cuckoo algorithm; Step S3: Using the Cuckoo algorithm, output the observation scheme for observing dense cluster targets using multiple agile satellites; In step S1, there are Ns agile satellites and Nt observation targets. The planning evaluation function F is: ， Among them, Toc i This represents the time spent observing the i-th target. The constraints include at least agile satellite attitude maneuver constraints, visible window execution constraints, agile satellite imaging observation time constraints, solar elevation angle constraints, and mission condition constraints; wherein, the agile satellite attitude maneuver constraint is expressed as: if the same satellite observes two adjacent targets, the actual observation start time between targets i and j and The interval is greater than or equal to the sum of the maneuvering time and imaging time between the two observed targets; the visible window execution constraint is expressed as: the satellite only operates within the visible execution window of the observed target point. Let i be the observation start time for target i and satellite j on the k-th orbit. To determine the end time of observation of target i and satellite j in the kth orbit, the visible time window of target i and satellite j within the kth orbit is defined. Let the visible start time be for target i and satellite j. Let the visible end time be for target i and satellite j, then we have: The time constraint for agile satellite imaging observations is expressed as: , The solar altitude angle constraint is that the solar altitude angle should be greater than the minimum required for Earth observation when the visible light camera images the ground. The expression is: ,in, This is the minimum solar altitude angle; The mission constraint is that a satellite performs only one mission at a time. The mission constraint expression is as follows: ; In step S2, the visible task sequence is encoded based on the Cuckoo algorithm, specifically as follows: ， in, Let Toc represent a set of observation schemes for N observation targets obtained for the i-th bird's nest in generation t. N This represents the observation time of the Nth objective obtained from the planning evaluation function.

2. The method according to claim 1, characterized in that, Step S3 specifically includes: Step S301: Update the cuckoo nest location by adjusting Lévy's flight stride and discovery probability to optimize the task sequence; Step S302: Dynamically adjust bird nests through learning strategies, discarding inferior bird nests and generating new bird nests; Step S303: Determine whether the termination condition is met and output the best bird's nest; In this context, any egg in a cuckoo's nest represents a task scheduling scheme, and any cuckoo has exactly one egg.

3. The method according to claim 2, characterized in that, In step S301, when the i-th cuckoo generates a new solution, Lévy's flight is executed as follows: In the formula, α > 0 represents the step size control variable, Num represents the number of hosts, and t represents the current generation. This represents the i-th bird's nest in generation t. This represents point-to-point multiplication. express Flight path, among which, ; λ is Flight parameters are the average or expected values ​​of events occurring within a unit interval. The calculation formula is: ， in, and v are the parameters of the normal distribution. and It is the standard deviation of the normal distribution. in, , , Let g denote the gamma function. Lévy flight provides a random walk, with the random step size derived from the Lévy distribution function.

4. The method according to claim 3, characterized in that, In step S302, the bird's nest moves towards the global optimal position, and the global optimal value in the current iteration is always greater than or at least equal to the global optimal value in the previous iteration, F(x g (t)) and F(x g (t -1)) represent the optimal solutions for the bird's nest found in the current iteration and the last iteration, respectively, including: The nest evolution factor of the learning cuckoo is defined by the following formula: The nest evolution factor utilizes historical information from the algorithm and reflects the search speed of Levy's flight. in, As md decreases, the current optimal solution for the bird's nest will approach the optimal solution. The nest clustering factor of learning-type cuckoos is defined by the following formula: , Where Mt represents the average of the optimal locations of all bird nests, and the formula is: F(xi(t)) represents the function value of the optimal solution F(xg(t)) in the current iteration that is better than that of each bird's nest.

5. The method according to claim 4, characterized in that, The step size control value and the detection probability are updated using the following formula: in, For the initial step size and The initial probability of finding a bird egg. By learning from historical information and taking the following actions: ； in, As the step size efficiency factor, To discover the probabilistic power factor, and These represent the effects of the nest evolution factor and the nest aggregation factor on the step size, respectively. The impact effect and These represent the influence of the nest evolution factor and the nest clustering factor on the discovery probability p, respectively. a The impact and effect.

6. A device, characterized in that, include: One or more processors, one or more memories, and one or more computer programs; wherein the processor is connected to the memory, and the one or more computer programs are stored in the memory, and when the device is running, the processor executes the one or more computer programs stored in the memory to cause the device to perform a planning method for multi-agile satellite observation dense cluster targets as described in any one of claims 1 to 5.

7. A computer-readable storage medium, characterized in that, Used to store computer instructions, which, when executed by a processor, implement the planning method for a dense cluster of multi-agile satellite observation targets as described in any one of claims 1 to 5.

Images