A multi-satellite planning method, device and medium based on a deep learning neural network

Abstract

The present invention discloses a multi-satellite planning method, device and medium based on a deep learning neural network. The method includes: obtaining historical planning data of multi-satellite planning, and after preprocessing, obtaining resource information of the deep learning neural network for multi-satellite planning; determining a constraint function according to the constraint information to be considered in multi-satellite planning as the model constraint condition of the deep learning neural network; demarcating a flag bit, and selecting resource information feature values as the input of the deep learning neural network according to the observed target task; determining an objective function according to the satisfaction degree of the observation task and the utilization rate of satellite payloads in multi-satellite planning, and training the deep learning neural network; and performing multi-satellite planning by using the trained deep learning neural network. The present invention provides a new multi-satellite planning method based on a deep learning neural network, which can make the multi-satellite planning task have better redundancy, robustness and task execution ability, and complete the user's multi-satellite planning task requirements faster and more accurately.

Description

Technical Field

[0001] The present invention relates to the field of establishing a deep learning neural network for multi-satellite planning in satellite imaging work, and in particular to a multi-satellite planning method, device and medium based on a deep learning neural network. Background Art

[0002] With the vigorous development of satellite technology, the mature application of multi-satellite technology, and the rise of commercial aerospace companies, the types of imaging satellites have gradually become richer and the scale has gradually grown, which has not only led to a significant increase in users' demand for satellite imaging tasks, but also made the classification of imaging tasks more complex and detailed. Especially in some special cases, users have higher requirements for the accuracy, clarity, timeliness and other aspects of imaging tasks. Therefore, in the process of satellite imaging, the limitation of independent work of a single satellite has been broken away, forming a comprehensive system of coordinated satellites of various types. In addition, due to the increasing requirements for high-resolution imaging and the limited number of satellites within a short transit range, the use of a single satellite can hardly meet the complex task requirements of regional image acquisition, while multi-satellite collaboration has better redundancy, robustness and mission execution capabilities. Therefore, it is necessary to use multiple satellites with the same sensor type and similar spatial resolution for collaborative imaging to ensure faster completion of user mission requirements.

[0003] Compared with traditional networking satellites and single satellite platforms, multi-satellite collaboration has its own characteristics: First, the functions of multi-satellite collaboration units are becoming increasingly complex, the degree of space mission coupling is higher, and the requirements for task coordination and allocation are higher. The satellite needs to have autonomous task planning capabilities. Secondly, multi-satellite multi-task units make it difficult to expose potential conflicts between satellites or tasks. The mission planning satellite needs to process a large amount of satellite interaction data. In the absence of multi-satellite simulation equipment, traditional simulation test methods are difficult to discover deep-seated conflicts and the impact on multi-satellite units. Thirdly, autonomous mission planning, as the core of completing space missions and tapping the potential of multi-satellite collaboration units, needs to have high reliability and security. If the algorithm function of the multi-satellite collaboration unit is not fully tested, it will cause the satellite function test to fail at best, and cause serious damage to the products and equipment on the satellite at worst. Finally, there are not many simulation verification methods for multi-satellite collaboration. Existing satellite simulation tests are mostly conducted on formations or single satellites. There is insufficient simulation capability to support multi-satellite units with heterogeneous functions, the need for on-orbit collaboration, and frequent information interaction. Against the backdrop of growing demand for satellite imaging and limited capabilities of single satellites, multi-satellite collaboration is an important means to meet diversified observation needs. However, due to different priorities of observation needs and different capability constraints of satellites, there is a problem of conflicts in the use of multiple resources when performing tasks.

[0004] The patent document with the publication number CN115016910A discloses a multi-satellite collaborative mission planning method and device for multi-observation tasks. This application simplifies the multi-objective planning problem for multi-observation tasks into multiple serial single-objective planning problems. The model can be solved quickly and has strong interpretability. However, it does not use a neural grid to construct a multi-satellite planning model.

[0005] The patent document with the publication number CN115271313A. The present invention relates to a multi-satellite collaborative mission planning method and device based on the conflict degree of observation requirements. This application improves the satisfaction rate of observation requirements and enhances the utilization efficiency of satellite resources. However, this application also does not use a neural grid to construct a multi-satellite planning model. Summary of the Invention

[0006] In view of the above problems, the purpose of the present invention is to provide a multi-satellite planning method, device and medium based on a deep learning neural network, so as to solve the conflict problem when multi-satellites execute tasks, and further shorten the multi-satellite planning time and improve the task satisfaction.

[0007] Embodiments of the present invention provide a multi-satellite planning method, device and medium based on a deep learning neural network.

[0008] First aspect: A multi-satellite planning method based on a deep learning neural network, including:

[0009] S1. Obtain the historical planning data of multi-satellite planning. After preprocessing, obtain the resource information of the multi-satellite planning deep learning neural network;

[0010] S2. Determine the constraint function according to the constraint information to be considered in multi-satellite planning, and use it as the model constraint condition of the deep learning neural network;

[0011] S3. Designate a flag bit, and select the resource information feature value according to the observed target task as the input of the deep learning neural network;

[0012] S4. Determine the objective function according to the observation task satisfaction and satellite payload utilization rate of multi-satellite planning, and train the deep learning neural network;

[0013] S5. Use the trained deep learning neural network for multi-satellite planning.

[0014] The resource information includes: satellite payload set, observed target set, single-satellite task set, and visible arc segment set of the satellite for the target.

[0015] Optionally: The:

[0016] The satellite payload set is defined as ,

[0017] where, , the total number of satellites is , , is the identifier of satellite . is the status of satellite . is the storage of satellite . is the payload type of satellite n, is the priority of satellite ;

[0018] The set of observed targets is defined as ,

[0019] where , the total number of targets is , , is the identifier of mission . is the status of target . is the type of target . is the priority of target ;

[0020] The set of single-satellite missions is defined as ,

[0021] where , the total number of missions is , , is the identifier of mission . is the satellite to which mission belongs, is the type of mission . is the start time of mission . is the end time of mission . is the execution status of mission . is the priority of mission ;

[0022] The set of visible arc segments of a satellite with respect to a target is defined as ,

[0023] , the total number of arc segments is , , is the identifier of arc segment . is the arc segment The corresponding satellite, is an arc segment corresponding to the target, is an arc segment for the start time, is an arc segment for the duration, is an arc segment for the end time, is an arc segment for the usage status, is an arc segment for the priority.

[0024] Optionally: The multi-satellite planning in S2 needs to consider constraint information including:

[0025] The failure probability of the satellite payload in multi-satellite planning is 0;

[0026] A single satellite payload observes only one target at the same time;

[0027] Each observation task is executed only once and only one arc segment is selected for execution;

[0028] The time interval between the executions of different tasks must be greater than or equal to the switching time of the device;

[0029] The effective execution time of the observation task should be greater than or equal to the sum of the task duration and the payload switching time.

[0030] Optionally: According to the constraint information, a constraint function is constructed, and the formula is expressed as:

[0031] A single satellite payload observes only one target at the same time:

[0032]

[0033] Each observation task is executed only once and only one arc segment is selected for execution:

[0034]

[0035] The time interval between the executions of different tasks must be greater than or equal to the switching time of the device :

[0036]

[0037] The effective execution time of the observation task should be greater than or equal to the sum of the task duration and the payload switching time:

[0038]

[0039] Among them, is a constant, indicating the task Whether it is executed in the satellite payload of the arc segment, is the end time of the effective execution of the observation task, is the start time of the effective execution of the observation task, is a constant representing the interval of the observation task, represents the effective execution time of the task, represents the duration of the task, represents the conversion time of the satellite payload.

[0040] Optionally: The satisfaction of the multi-satellite planned observation task is expressed by the formula:

[0041]

[0042] The utilization rate of the multi-satellite planned satellite payload is expressed by the formula:

[0043]

[0044] Indicates whether the task is successfully executed, 1 for success and 0 for failure, Indicates whether the visible arc segment of the satellite payload is allocated, 1 for allocated and 0 for unallocated, is the visible arc segment of the multi-satellite planned satellite payload.

[0045] Optionally: Determine the objective function based on the satisfaction of the multi-satellite planned observation task and the utilization rate of the satellite payload, expressed by the formula:

[0046]

[0047] Among them, is a constant, is the satisfaction of the observation task, is the utilization rate of the satellite payload.

[0048] Optionally: The flag bits in the S3 include:

[0049] Set the arc segment constraint flag bit according to the observed target task , when the visible arc segment duration of the satellite for the target is greater than or equal to the task duration it is 1, and when it is less it is 0;

[0050] Weighted sum of the priorities of the satellite, target, arc segment, and task according to the observed target task, and record this value as the priority flag bit ;

[0051] Set a conflict flag according to the task of the observed target 。

[0052] Second aspect: An electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, it implements the steps of the method provided in the first aspect.

[0053] Third aspect: A non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements the steps of the method provided in the first aspect.

[0054] Advantages of the present invention:

[0055] The present invention utilizes the historical planning data of multi-satellite planning to form the resource information of the deep learning neural network for multi-satellite planning. According to the constraint information that needs to be considered in multi-satellite planning, it is used as the model constraint condition of the deep learning neural network. According to the task of the observed target, the flag bit is delimited to select the resource information eigenvalue as the input of the deep learning neural network. According to the satisfaction degree of the observation task and the satellite payload utilization rate of multi-satellite planning, as the objective function, the deep learning neural network is trained; finally, the optimal weight matrix and the offset vector of the multi-satellite planning network model are obtained. Then, the trained deep learning neural network is used for multi-satellite planning, providing a new multi-satellite planning method based on the deep learning neural network, which can make the multi-satellite planning task have better redundancy, robustness and task execution ability, and complete the user's multi-satellite planning task requirements faster and more accurately. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 is a schematic flow chart of a multi-satellite planning method based on a deep learning neural network of the present invention;

[0057] Figure 2 is a schematic diagram of the flow principle of a multi-satellite planning method based on a deep learning neural network of the present invention;

[0058] Figure 3 is a schematic diagram of the structure of the electronic device of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0059] The embodiments of the present invention will be described in detail below. The examples of the embodiments are shown in the drawings, where the same or similar symbols represent the same or similar elements or elements with the same or similar functions from beginning to end. The embodiments described below by referring to the drawings are exemplary and are only used to explain the present invention and should not be construed as a limitation of the present invention.

[0060] The functions of multi-satellite cooperation units are gradually becoming more complex, the coupling degree of space missions is higher, and the requirements for mission cooperation and allocation are also higher. In the absence of multi-satellite simulation equipment, it is difficult for traditional simulation testing methods to discover deep-seated conflicts and the impacts on multi-satellite units. The existing simulation verification means for multi-satellite cooperation are not rich enough. Existing satellite simulation tests are mostly carried out for formations or single satellites, and the simulation capabilities for multi-satellite units with heterogeneous functions, requiring on-orbit cooperation, and frequent information interaction are insufficient.

[0061] To address the above problems, the present invention provides a multi-satellite planning method based on a deep learning neural network. Figure 1 The following is a schematic flowchart of the multi-satellite planning method based on a deep learning neural network provided by an embodiment of the present invention. The method includes:

[0062] S1. Obtain the historical planning data of multi-satellite planning. After preprocessing, obtain the resource information of the deep learning neural network for multi-satellite planning.

[0063] The main principle of a deep learning neural network is to achieve complex function approximation between input and output through self-learning of the features of a large number of samples. Specifically, a deep learning neural network constructs a model with multiple network layers and a large amount of training data to learn more useful features, thereby ultimately improving the accuracy of classification or prediction, avoiding the cumbersome process of manually extracting features, and greatly improving the efficiency and accuracy of data processing and analysis. Deep learning neural networks have a wide range of applications, and the structure of the deep learning neural network is not specifically introduced here.

[0064] After obtaining the multi-satellite planning historical data and results, delete the tasks with failed multi-satellite planning, and construct a set of data with a 100% success rate of multi-satellite planning as the resource information. The collected resource information mainly includes: satellite payload set, observed target set, single-satellite task set, and the visible arc segment set of the satellite for the target.

[0065] Among them, the satellite payload set is defined as ,

[0066] where , the total number of satellites is , , is the identifier of satellite , is the state of satellite , is the storage of satellite , is the payload type of satellite n, is the satellite priority;

[0067] The observed target set is defined as ,

[0068] Among them, , the total number of targets is , , is the identifier of task , is the status of target , is the type of target , is the priority of target ;

[0069] The set of single-satellite tasks is defined as ,

[0070] Among them, , the total number of tasks is , , is the identifier of task , is the satellite to which task belongs, is the type of task , is the start time of task , is the end time of task , is the execution status of task , is the priority of task ;

[0071] The set of visible arc segments of a satellite with respect to a target is defined as ,

[0072] , the total number of arc segments is , , is the identifier of arc segment , is the satellite corresponding to arc segment , is the target corresponding to arc segment , is the start time of arc segment , is the duration of arc segment , is the end time of arc segment , is the usage status of arc segment , is the priority of arc segment .

[0073] S2. Determine the constraint function based on the constraint information to be considered in multi-satellite planning, and use it as the model constraint condition of the deep learning neural network.

[0074] The construction of the deep learning neural network model needs to meet the objective constraint conditions of multi-satellite planning. The constraint information to be considered in multi-satellite planning mainly includes:

[0075] The failure probability of the satellite payload in multi-satellite planning is 0; a single satellite payload observes only one target at the same time; each observation task is executed only once and only one arc segment is selected for execution; the time interval between the executions of different tasks must be greater than or equal to the switching time of the device; the effective execution time of the observation task should be greater than or equal to the sum of the task duration and the payload switching time, etc.

[0076] The specific constraint information can be expressed by the constraint function as:

[0077] A single satellite payload observes only one target at the same time:

[0078]

[0079] Each observation task is executed only once and only one arc segment is selected for execution:

[0080]

[0081] The time interval between the executions of different tasks must be greater than or equal to the switching time of the device :

[0082]

[0083] The effective execution time of the observation task should be greater than or equal to the sum of the task duration and the payload switching time:

[0084]

[0085] Among them, is a constant, indicating whether the task is executed in the arc segment of the satellite payload , is the effective execution end time of the observation task, is the effective execution start time of the observation task, is a constant, indicating the interval of the observation task, represents the effective execution time of the task, represents the duration of the task, represents the switching time of the satellite payload.

[0086] By constraining the deep learning neural network with the above-mentioned constraint information, the deep learning neural network is enabled to meet the conditions for multi-satellite planning and coordination, and has better redundancy, robustness, and task execution capabilities.

[0087] S3. Designate flag bits, and select the resource information eigenvalue as the input of the deep learning neural network according to the task of the observed target.

[0088] After the resource information and constraint information involved in the multi-satellite planning task are defined, further select the resource information that has a significant impact on the multi-satellite planning result as the eigenvalue input to the deep learning neural network, and train the deep learning neural network; specifically, the resource information eigenvalue can be obtained by setting flag bits, and the flag bits include:

[0089] Set the arc segment constraint flag bit , the quality of the visible arc segment of the satellite for the target has a great impact on the planning result. According to the task of the observed target, when the visible arc segment duration of the satellite for the target is greater than or equal to the task duration it is 1, and when it is less it is 0. The resource information can be screened and limited according to the task of the observed target, which is more conducive to the training of the deep learning neural network.

[0090] Similarly, the priority of the same observation task also affects the quality of the multi-satellite planning task result. According to the task of the observed target, the priorities of the satellite, arc segment, and task can be weighted and summed, and this value is recorded as the priority flag bit , and the priority of the observation task is distinguished according to the flag bit .

[0091] During the multi-satellite planning process, conflict problems may also occur. The conflict flag bit can be set according to the task of the observed target , which is 1 when there is a conflict and 0 otherwise.

[0092] In addition, the utilization rate of the satellite payload can also be considered during the planning process to prevent the satellite payload resources from working stagnantly, and the load balance flag bit is set , which is 1 when the payload utilization rate is greater than or equal to 80% and 0 otherwise .

[0093] S4. Determine the objective function according to the satisfaction degree of the multi-satellite planning observation task and the satellite payload utilization rate, and train the deep learning neural network;

[0094] Among them, the satisfaction degree of the multi-satellite planning observation task is expressed by the formula:

[0095]

[0096] The satellite payload utilization rate for multi-satellite planning is expressed by the formula:

[0097]

[0098] where indicates whether the task is successfully executed, with a value of 1 for success and 0 for failure. indicates whether the visible arc segment of the satellite payload is allocated, with a value of 1 for allocation and 0 for non-allocation. is the visible arc segment of the satellite payload for multi-satellite planning.

[0099] Based on the observation task satisfaction and satellite payload utilization rate for multi-satellite planning, the objective function formula is expressed as:

[0100]

[0101] where is a constant, is the observation task satisfaction, is the satellite payload utilization rate.

[0102] As Figure 2 shown, the deep learning neural network is trained according to the objective function until the objective function is lower than the set threshold, and it can be considered that the training of the deep learning neural network is completed.

[0103] S5. Use the trained deep learning neural network for multi-satellite planning.

[0104] As Figure 2 shown, after the deep learning neural network is trained, it can be verified and tested. Using the test set formed by the resource information, the trained deep learning neural network is tested. After testing, when the deep learning neural network does not meet the usage requirements, retraining is performed. After the deep learning neural network meets the usage requirements, it can be used for multi-satellite planning tasks.

[0105] After obtaining the historical data of multi-satellite planning in the present invention, the task data of failed planning is deleted, a set of resource information data with a planning success rate of 100% is constructed, and then according to the observed target task, a flag bit is demarcated. The resource information feature value is selected as the input of the deep learning neural network according to the flag bit. Then, according to the observation task constraint conditions of multi-satellite planning, in order to meet the observation task satisfaction and satellite payload utilization rate of multi-satellite planning, the objective function is determined, and the deep learning neural network is trained. Finally, the optimal weight matrix and offset vector of the deep learning neural network model are used for multi-satellite planning.

[0106] The present invention also provides an electronic device. Figure 3The structural schematic diagram of the electronic device provided by the embodiment of the present invention is as follows. Figure 3 As shown in the figure, the electronic device may include: a processor, a communication interface, a memory, and a communication bus. Among them, the processor, the communication interface, and the memory complete mutual communication through the communication bus. The processor can call the logical instructions in the memory to execute the following method:

[0107] S1. Obtain the historical planning data of multi-satellite planning. After preprocessing, obtain the resource information of the multi-satellite planning deep learning neural network;

[0108] S2. Determine the constraint function according to the constraint information to be considered in multi-satellite planning, and use it as the model constraint condition of the deep learning neural network;

[0109] S3. Define a flag bit, and select the resource information eigenvalue according to the observed target task as the input of the deep learning neural network;

[0110] S4. Determine the objective function according to the satisfaction degree of the observation task of multi-satellite planning and the utilization rate of satellite payloads, and train the deep learning neural network;

[0111] S5. Use the trained deep learning neural network for multi-satellite planning.

[0112] In addition, when the logical instructions in the above-mentioned memory are implemented in the form of software functional units and sold or used as an independent product, they can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. And the aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical discs that can store program codes.

[0113] The embodiment of the present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it is used to execute the methods provided in the above-mentioned various embodiments, for example, including:

[0114] S1. Obtain the historical planning data of multi-satellite planning. After preprocessing, obtain the resource information of the multi-satellite planning deep learning neural network;

[0115] S2. Determine a constraint function based on the constraint information to be considered in multi-satellite planning as the model constraint condition of the deep learning neural network;

[0116] S4. Define a flag bit, and select the resource information eigenvalue as the input of the deep learning neural network according to the task of the observed target;

[0117] S7. Determine an objective function based on the satisfaction degree of the observation task and the satellite payload utilization rate in multi-satellite planning, and train the deep learning neural network;

[0118] S5. Use the trained deep learning neural network for multi-satellite planning.

[0119] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art can understand and implement it without creative work.

[0120] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the essence of the above technical solution or the part that contributes to the prior art can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., including several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0121] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A multi-star planning method based on deep learning neural network, characterized in that: include: S1. Obtain historical planning data of multi-satellite planning, and after preprocessing, obtain resource information of multi-satellite planning deep learning neural network; S2. Determine the constraint function based on the constraint information that needs to be considered in multi-satellite planning as the model constraint condition of the deep learning neural network; S3, define the flag bit, and select the resource information feature value as the input of the deep learning neural network according to the observed target task; S4. Determine the objective function and train the deep learning neural network based on the observation mission satisfaction and satellite payload utilization of multi-satellite planning; S5. Use the trained deep learning neural network for multi-star planning; The resource information includes: a satellite payload set, an observed target set, a single satellite mission set, and a satellite visible arc set for the target; The multi-star planning needs to consider the following constraints: The probability of satellite payload failure in multi-satellite planning is 0; A single satellite payload observes only one target at a time; Each observation task is executed only once and only one arc segment is selected for execution; The time interval between executing different tasks must be greater than or equal to the switching time of the device; The effective execution time of the observation task must be greater than or equal to the sum of the task duration and the load switching time; The objective function is expressed as: max[αTR+(1-α)SR] Among them, α is a (0,1) constant, TR is the observation mission satisfaction, and SR is the satellite payload utilization rate; The flag bits in S3 include: According to the observed target task, set the arc constraint flag A r , the satellite's visible arc duration for the target is greater than or equal to the mission duration A r is 1, less than A r is 0; According to the observed target mission, the priority of the satellite, target, arc segment and mission is weighted and summed, and this value is recorded as the priority flag bit P r ; According to the observed target task, set the conflict flag C r .

2. A multi-star planning method based on deep learning neural network according to claim 1, characterized in that: Said: The satellite payload set is defined as S = {s1, s2, s3, …, s n ,…,s N }, in, The total number of satellites is N, 1≤n≤N, is the identifier of satellite n, is the state of satellite n, is the storage of satellite n, is the payload type of satellite n, is the priority of satellite n; The set of observed targets is defined as O = {o1, o2, o3, …, o m ,…,o M }, in, The total number of targets is M, 1≤m≤M, is the identifier of task m, is the state of target m, is the type of target m, is the priority of target m; The single-star mission set is defined as T = {t1, t2, t3, …, t l ,…,t L }, in, The total number of tasks is L, 1≤l≤L, is the identifier of task l, is the satellite to which mission l belongs, is the type of task l, is the start time of task l, is the end time of task l, is the execution status of task l, is the priority of task l; The set of visible arcs of the satellite to the target is defined as A = {a1, a2, a3, ..., a k ,…,a K }, The total number of arc segments is K, 1≤k≤K, is the identifier of arc k, is the satellite corresponding to arc segment k, is the target corresponding to arc segment k, is the start time of arc segment k, is the duration of arc segment k, is the end time of arc segment k, is the usage status of arc segment k, is the priority of arc k.

3. A multi-star planning method based on deep learning neural network according to claim 2, characterized in that: Based on the constraint information, the constraint function is constructed, and the formula is expressed as: A single satellite payload observes only one target at a time: Each observation task is executed only once and only one arc segment is selected for execution: The time interval between executing different tasks must be greater than or equal to the device switching time β: The effective execution time of the observation task must be greater than or equal to the sum of the task duration and the load switching time: in, is a (0,1) constant, indicating whether task l is executed in arc segment k of satellite payload n. is the effective execution end time of the observation task, is the effective execution start time of the observation task, β is a constant, indicating the interval of the observation task, Indicates the effective execution time of the task. represents the duration of the task, t transform Indicates the switching time of the satellite payload.

4. The multi-star planning method based on deep learning neural network according to claim 3 is characterized in that: The satisfaction degree of multi-satellite planned observation mission is expressed as: The satellite payload utilization rate of multi-satellite planning is expressed as follows: x l Indicates whether the task is successfully executed, 1 if successful, 0 if failed. Indicates whether the visible arc of the satellite payload is allocated, 1 if allocated, and 0 if not allocated. total Plan satellite payload visible arcs for multiple satellites.

5. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the steps of a multi-star planning method based on a deep learning neural network are implemented as described in any one of claims 1 to 4.

6. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of a multi-star planning method based on a deep learning neural network as described in any one of claims 1 to 4 are implemented.

Images