Multi-satellite multi-phase observation task planning method and system based on genetic iterative search

Abstract

The application provides a multi-satellite multi-stage observation task planning method and system based on genetic iteration search. A set of to-be-observed tasks and a set of observation time windows of satellites are acquired, and a set of cost coefficients is constructed. The set of cost coefficients includes a plurality of cost coefficient combinations, and each cost coefficient combination includes cost coefficients of all to-be-observed tasks. Based on the set of to-be-observed tasks and the set of observation time windows, an initial to-be-observed task planning scheme corresponding to each cost coefficient combination in the set of cost coefficients is acquired, and a scheme profit of the initial to-be-observed task planning scheme is calculated. The set of cost coefficients is updated based on the scheme profit, and a to-be-observed task planning scheme is acquired based on the updated set of cost coefficients. The effect and reliability of the planning scheme are improved.

Description

Technical Field

[0001] This invention relates to the field of satellite technology, and specifically to a multi-satellite, multi-stage observation mission planning method and system based on genetic iterative search. Background Technology

[0002] Earth observation satellites are an important type of spacecraft. With their widespread application in various fields, user needs are becoming increasingly complex and diverse, generating numerous complex Earth observation requirements, which can be termed multi-satellite, multi-stage observation missions. Executing such missions requires using multiple designated types of satellites, following a specific sequence, to image the same ground target multiple times within a scheduled period. Therefore, how to plan multi-satellite, multi-stage observation missions has become a research hotspot in this field.

[0003] For multiple multi-satellite, multi-stage observation missions, different missions will compete for limited satellite resources. The missions planned earlier will occupy some satellite resources, thus affecting the missions planned later. When satellites are observing, the cost of using the observation time window directly affects the planning scheme of each mission. Therefore, it is necessary to plan multi-satellite, multi-stage observation missions in conjunction with the cost of using the observation time window.

[0004] Existing mission planning methods generally employ manual planning, where technicians plan each observation mission individually. However, manual planning methods do not consider the impact of mission sequencing and the cost of using observation time windows on the final overall planning scheme, resulting in poor overall mission planning performance and low reliability. Summary of the Invention

[0005] (a) Technical problems to be solved

[0006] To address the shortcomings of existing technologies, this invention provides a multi-star, multi-stage observation mission planning method and system based on genetic iterative search, which solves the technical problems of poor overall mission planning performance and low reliability in existing technologies.

[0007] (II) Technical Solution

[0008] To achieve the above objectives, the present invention provides the following technical solution:

[0009] This invention provides a multi-star, multi-stage observation mission planning method based on genetic iterative search to solve its technical problem. The planning method is executed by a computer and includes the following steps:

[0010] Obtain the set of tasks to be observed and the set of observation time windows for satellites; construct a set of cost coefficients, which includes several cost coefficient combinations, each cost coefficient combination including the cost coefficient of the task to be observed;

[0011] Based on the set of tasks to be observed and the set of observation time windows, obtain the initial task planning scheme corresponding to each cost coefficient combination in the cost coefficient set, and calculate the scheme profit of the initial task planning scheme.

[0012] Update the cost coefficient set based on the profit of the proposed scheme;

[0013] The planning scheme for the task to be observed is obtained based on the updated set of cost coefficients.

[0014] Preferably, the step of obtaining the initial observation task planning scheme corresponding to each cost coefficient combination in the cost coefficient set includes:

[0015] Obtain the task benefits of all tasks to be observed and the usage costs of all observation time windows, and sort the tasks to be observed in descending order of the task benefits;

[0016] Based on the sorting results, and using the observation time window set and the usage cost, the target observation time window set for all tasks to be observed is obtained sequentially.

[0017] Based on the target observation time window set of all tasks to be observed, obtain the initial task planning scheme corresponding to each combination of cost coefficients.

[0018] Preferably, the step of sequentially obtaining the target observation time window set for all tasks to be observed includes:

[0019] Within the set of observation time windows, obtain the first set of observation time windows corresponding to the first task to be observed; obtain the observation benefits, observation frequency, observation satellite type, and observation satellite order of the first task to be observed;

[0020] Calculate the product of the usage cost of the first observation time window and the cost coefficient of the first task to be observed, and use it as the update cost of the first observation time window;

[0021] Based on the first observation time window, the type of the observed satellite, and the order of the observed satellites, a directed graph of the observation time window path corresponding to the first task to be observed is obtained.

[0022] The target observation time window set of the first task to be observed is obtained based on the directed graph of the observation time window path.

[0023] Continue to obtain the target observation time window set for the next task to be observed, until the target observation time window set for all tasks to be observed is obtained.

[0024] Preferably, obtaining the directed graph of the observation time window path corresponding to the first task to be observed includes:

[0025] Establish a start point and an end point, and based on the observation frequency, set the observation phase for the first task to be observed;

[0026] Based on the observed satellite type and the observed satellite order, the first observation time window is added to the corresponding observation stage, and a virtual observation time window is established in each observation stage;

[0027] Based on preset connection conditions, the observation time windows in different observation stages are connected by directed arcs to obtain a directed graph of observation time window paths.

[0028] Preferably, obtaining the target observation time window set of the first task to be observed based on the directed graph of the observation time window path includes:

[0029] Based on the update cost, calculate the time window profit of the first observation time window, and set the time window profits of the start point, the end point and the virtual observation time window to 0;

[0030] The parent observation time window of each observation time window in the first observation stage is set as the starting point, and the path profit of each observation time window in the first observation stage is obtained; the path profit is the sum of the time window profit of the observation time window and the path profit of the parent observation time window; the path profit of the starting point is 0.

[0031] In the second observation phase, obtain the node with the largest path profit among the starting nodes of all directed arcs corresponding to each observation time window, and use it as the parent observation time window for each observation time window; obtain the path profit for each observation time window in the second observation phase.

[0032] Based on the path profit of each observation time window in the second observation stage, obtain the parent observation time window and path profit of each observation time window in the next observation stage, until the parent observation time window and path profit of each observation time window in the last observation stage are obtained.

[0033] The observation time window with the highest path profit in the last observation stage is determined as the target observation time window; all parent observation time windows corresponding to the target observation time window are obtained, and the parent observation time window path of the target observation time window is determined based on the target observation time window, all parent observation time windows and the endpoint;

[0034] All first observation time windows in the parent observation time window path are determined as the target observation time window set for the first task to be observed.

[0035] Preferably, obtaining the target observation time window set for the next task to be observed includes:

[0036] Obtain the incompatible observation time windows of all observation time windows in the target observation time window set of the first observation task, wherein the incompatible observation time windows are observation time windows that cannot be executed simultaneously.

[0037] The target observation time window set of the first observation task and the incompatible observation time window are deleted from the observation time window set to update the observation time window set;

[0038] Based on the updated set of observation time windows, obtain the target set of observation time windows for the next task to be observed.

[0039] Preferably, calculating the scheme profit of the initial observation task planning scheme includes:

[0040]

[0041] in:

[0042] F represents the profit of the project;

[0043] N T K represents the number of tasks to be observed. i n represents the observation frequency of the i-th observation task; i p represents the actual number of images for the i-th observation task; i This represents the observation reward for the i-th task to be observed;

[0044] D bl This represents the initial planning scheme for the observation task; w h c represents the observation time window in the initial planning scheme for the task to be observed; h Indicates the observation time window w h The cost of using it.

[0045] Preferably, updating the cost coefficient set based on the profit of the proposed scheme includes:

[0046] A selection operation is performed on the cost coefficient set. The selection operation is as follows: sort the initial task planning schemes corresponding to each cost coefficient combination according to the order of the scheme profit from large to small, and delete the cost coefficient combinations corresponding to the lower-ranked initial task planning schemes from the cost coefficient set.

[0047] A crossover operation is performed on the cost coefficient set. The number of crossover operations is a preset first threshold. The crossover operation is as follows: two cost coefficient combinations are randomly copied, and the cost coefficient of any one of the observed tasks in the two cost coefficient combinations is swapped. The two new cost coefficient combinations are then added to the cost coefficient set.

[0048] A mutation operation is performed on the cost coefficient set. The number of mutation operations is a preset second threshold. The mutation operation is as follows: a cost coefficient combination is randomly copied, and the cost coefficient of any one of the observed tasks in the cost coefficient combination is updated to any value between 0 and 1. The new cost coefficient combination is then added to the cost coefficient set.

[0049] Preferably, obtaining the planning scheme for the observed task based on the updated cost coefficient combination includes:

[0050] Set the number of iterations;

[0051] Based on the updated set of cost coefficients, the step of obtaining the initial observation task planning scheme corresponding to each combination of cost coefficients in the set of cost coefficients is re-executed; iterative updates are performed until the number of iterations is reached.

[0052] Iterate through all the initial task planning schemes and select the one with the highest profit as the task planning scheme.

[0053] The present invention provides a multi-star, multi-stage observation mission planning system based on genetic iterative search to solve its technical problem, comprising:

[0054] The acquisition module is configured to acquire a set of tasks to be observed and a set of observation time windows for satellites; and to construct a set of cost coefficients, which includes several cost coefficient combinations, each of which includes the cost coefficient of the task to be observed.

[0055] The profit calculation module is configured to obtain the initial planning scheme for each combination of cost coefficients in the cost coefficient set based on the set of tasks to be observed and the set of observation time windows, and to calculate the scheme profit of the initial planning scheme for tasks to be observed.

[0056] The update module is configured to update the set of cost coefficients based on the profit of the proposed scheme;

[0057] The optimal strategy acquisition module is configured to obtain the planning scheme for the task to be observed based on the updated set of cost coefficients.

[0058] (III) Beneficial Effects

[0059] This invention provides a method and system for planning multi-star, multi-stage observation missions based on genetic iterative search. Compared with existing technologies, it has the following advantages:

[0060] This invention acquires a set of tasks to be observed and a set of satellite observation time windows, and constructs a set of cost coefficients. The cost coefficient set includes several cost coefficient combinations, each containing the cost coefficients for all tasks to be observed. Based on the set of tasks and observation time windows, an initial task planning scheme corresponding to each cost coefficient combination in the cost coefficient set is obtained, and the scheme profit of the initial task planning scheme is calculated. The cost coefficient set is updated based on the scheme profit; a task planning scheme is obtained based on the updated cost coefficient set. Each task can have a cost coefficient assigned to all observation time windows. This cost coefficient allows for the adjustment of the time window cost, thus planning each task. Based on the profit of all planning schemes, the cost coefficients can be updated to obtain a greater profit, resulting in a final task planning scheme, thereby improving the effectiveness and reliability of the planning scheme. Attached Figure Description

[0061] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0062] Figure 1 This is a flowchart of a multi-star, multi-stage observation mission planning method based on genetic iterative search provided in an embodiment of the present invention;

[0063] Figure 2 This is a schematic diagram of the directed graph of the observation time window path in an embodiment of the present invention. Detailed Implementation

[0064] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention are described clearly and completely. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0065] This application provides a multi-star, multi-stage observation mission planning method and system based on genetic iterative search, which solves the problems of poor overall mission planning performance and low reliability in existing technologies, and improves the effectiveness and reliability of the planning scheme.

[0066] The technical solution in this application is to solve the above-mentioned technical problems, and the general idea is as follows:

[0067] This invention provides an embodiment that acquires a set of tasks to be observed and a set of satellite observation time windows, and constructs a set of cost coefficients. The cost coefficient set includes several cost coefficient combinations, each containing the cost coefficients for all tasks to be observed. Based on the set of tasks and observation time windows, an initial task planning scheme corresponding to each cost coefficient combination in the cost coefficient set is obtained, and the scheme profit of the initial task planning scheme is calculated. The cost coefficient set is updated based on the scheme profit; a task planning scheme is then obtained based on the updated cost coefficient set. Each task to be observed can have a cost coefficient set for all observation time windows. This cost coefficient allows for the adjustment of the time window cost, thereby planning for each task. Based on the profit of all planning schemes, the cost coefficients can be updated to obtain a greater profit, resulting in a final task planning scheme, thus improving the effectiveness and reliability of the planning scheme.

[0068] To better understand the above technical solutions, the following will provide a detailed explanation of the technical solutions in conjunction with the accompanying drawings and specific implementation methods.

[0069] This invention provides a multi-star, multi-stage observation mission planning method based on genetic iterative search, which is executed by a computer, such as... Figure 1 As shown, Figure 1 This is a flowchart illustrating a multi-star, multi-stage observation mission planning method based on genetic iterative search, provided in an embodiment of the present invention. The method includes the following steps:

[0070] S1. Obtain the set of tasks to be observed and the set of observation time windows for satellites; construct a set of cost coefficients, which includes several cost coefficient combinations, each cost coefficient combination including the cost coefficient of the task to be observed;

[0071] S2. Based on the set of tasks to be observed and the set of observation time windows, obtain the initial task planning scheme corresponding to each cost coefficient combination in the cost coefficient set, and calculate the scheme profit of the initial task planning scheme.

[0072] S3. Update the cost coefficient set based on the profit of the proposed scheme;

[0073] S4. Obtain the planning scheme for the task to be observed based on the updated set of cost coefficients.

[0074] The following is a detailed analysis of each step.

[0075] In step S1, the set of tasks to be observed and the set of observation time windows of satellites are obtained, and the set of cost coefficients is constructed.

[0076] The set of tasks to be observed contains all tasks to be observed, consisting of N T It consists of multiple stages of observation tasks. The set of tasks to be observed can be represented as... Among them, t i This represents the i-th task, where 1 ≤ i ≤ N. T .

[0077] The embodiments of the present invention are configured with N S Several Earth observation satellites are used for observation missions; the Earth observation satellite set is as follows: s j This represents the j-th Earth observation satellite, and each Earth observation satellite carries an imaging sensor.

[0078] The observation time window set contains all observation time windows available from all satellites for each mission. The observation time window set is represented as... Among them, w h This represents the h-th time window, where 1 ≤ h ≤ N. w .

[0079] The cost coefficient set includes several cost coefficient combinations, and each cost coefficient combination includes the cost coefficients for all the tasks to be observed. Each cost coefficient combination contains N. T There are 1 cost coefficient, and each cost coefficient corresponds to an observation task. The cost coefficient can be a number between 0 and 1. Initially, each cost coefficient can be randomly generated.

[0080] The cost coefficient set can be set to N. B A combination of cost coefficients, expressed as b l This represents the l-th cost coefficient combination, where 1 ≤ l ≤ N. B .

[0081] Step S2 specifically includes the following steps:

[0082] S21. Based on the set of tasks to be observed and the set of observation time windows, obtain the initial planning scheme for each cost coefficient combination in the cost coefficient set. Each cost coefficient combination contains the cost coefficients for all tasks to be observed, and a planning scheme for each task to be observed can be obtained based on each cost coefficient. In this embodiment, the planning scheme for each task to be observed refers to the observation time window that the task ultimately matches when planning the task.

[0083] Under each cost coefficient combination, the planning schemes of all the tasks to be observed together constitute the initial planning scheme of the tasks to be observed corresponding to each cost coefficient combination.

[0084] Specifically, the following steps are included:

[0085] S211. Obtain the task revenue of all tasks to be observed in the set of tasks to be observed, and obtain the usage cost of all observation time windows in the set of observation time windows, and sort the tasks to be observed in descending order of the task revenue.

[0086] Specifically, the task benefits of the task to be observed and the usage cost of the observation time window can be considered as attributes of the task to be observed and the observation time window itself. Therefore, the task benefit data of each task to be observed and the usage cost data of each observation time window can be obtained directly.

[0087] All the tasks to be observed can be sorted in descending order of task reward. If there are multiple tasks with the same reward, they can be sorted randomly.

[0088] S212. Based on the sorting results, and using the observation time window set and the usage cost, sequentially obtain the target observation time window set for all tasks to be observed. The target observation time window set for each task to be observed is the planning scheme for that task, i.e., all observation time windows ultimately matched for each task. Specifically, this includes the following steps:

[0089] S2121. In the set of observation time windows, obtain the first set of observation time windows corresponding to the first task to be observed; obtain the observation benefits, observation frequency, observation satellite type and observation satellite order of the first task to be observed.

[0090] Specifically, each observation time window specifies which task it is used to observe. Therefore, all the first observation time windows corresponding to the first task can be filtered out to obtain the set of first observation time windows.

[0091] The observation benefits, observation frequency, observation satellite type, and observation satellite order of the first observation task are all attributes of the observation task itself, and therefore can be obtained directly.

[0092] S2122. Calculate the product of the usage cost of the first observation time window and the cost coefficient of the first task to be observed, and use it as the update cost of the first observation time window.

[0093] In this embodiment of the invention, the update cost is obtained by calculating the product of the usage cost and the cost coefficient. The update cost is then used to further filter the observation time window for each task to be observed, so that the update cost coefficient can be used to perturb the planning scheme to obtain a planning scheme with greater profit.

[0094] S2123. Based on the first observation time window, the type of the observation satellite, and the order of the observation satellites, obtain the directed graph of the observation time window path corresponding to the first task to be observed. Figure 2 This is a schematic diagram of the directed graph of the observation time window path in an embodiment of the present invention. Figure 2 As shown, the directed graph of observation time window paths includes: all observation time windows corresponding to the task to be observed in all observation stages, and the planning scheme paths that observation time windows can form.

[0095] Specifically, the following steps are included:

[0096] S21231. Establish a starting point and an ending point, and based on the observation frequency, set the observation phase of the first task to be observed.

[0097] Observation frequency refers to the frequency at which a task needs to be observed. Therefore, each task corresponds to its observation frequency and observation phases. For example, if the observation frequency of a task is 3, then it has 3 observation phases. Figure 2 Phases 1, 2, and 3. A starting point v is set for the observation task. s and the endpoint v e .

[0098] S21232. Based on the observation satellite type and the observation satellite sorting, the first observation time window is added to the corresponding observation stage, and a virtual observation time window is established in each observation stage.

[0099] The type of observation satellite refers to the type of satellites that need to be observed for the mission, and the order of observation satellites refers to the order in which all types of satellites are observed when observing the mission.

[0100] Based on the type and order of the observed satellites, all first observation time windows can be added sequentially to the corresponding observation phases. For example... Figure 2 In this study, Phase 1 involves multispectral satellites with two observation windows; Phase 2 involves visible light satellites with three observation windows; and Phase 3 involves synthetic aperture radar satellites with two observation windows. The first observation window is represented by a black node.

[0101] A virtual observation time window is established in each observation phase, represented by a white node.

[0102] S21233. Based on the preset connection conditions, the observation time windows in different observation stages are connected by directed arcs to obtain a directed graph of observation time window paths.

[0103] The default connection condition is: E min ≤E≤E max Where E represents the time interval between two observation windows, E min For the minimum threshold, E max This is the maximum threshold. Both thresholds can be values ​​set by the technicians.

[0104] For two adjacent observation phases, each phase can select one observation time window to connect, forming a path. However, if two observation time windows conflict, such as containing the same time, it means the two observation time windows cannot be performed simultaneously, and therefore a path cannot be formed. Therefore, based on the above connection conditions, all observation time windows that meet the conditions can be connected using directed arcs. Starting from the starting point and ending at the ending point constitutes a valid observation time window path. Thus, all observation time window paths are obtained.

[0105] S2124. Obtain the target observation time window set of the first task to be observed based on the directed graph of the observation time window path.

[0106] A directed graph of observation time windows may contain multiple observation time window paths, each with potentially different profits. To improve the effectiveness of the planning scheme, the path with the highest profit can be selected as the planning scheme. This involves the following steps:

[0107] S21241. Based on the update cost, calculate the time window profit of the first observation time window, and set the time window profits of the starting point, the ending point and the virtual observation time window to 0.

[0108] The profit of the first observation window can be calculated using the following formula:

[0109]

[0110] Among them, g m Indicates the observation time window w m Profit window;

[0111] p i Represents the task to be observed, t. i The observational gains, K i Indicates t i Observation frequency;

[0112] c m Indicates the observation time window w m The cost of updating.

[0113] For the start point, end point, and virtual observation time window, since they are not real observation time windows, their time window profit is set to 0.

[0114] Based on the observation time window profits of all nodes, the path with the highest profit can be obtained. The maximum profit can be determined sequentially from the starting point to each node in each observation stage, thus allowing for sequential filtering by observation stage to obtain the final path.

[0115] S21242. Set the parent observation time window of each observation time window in the first observation stage as the starting point, and obtain the path profit of each observation time window in the first observation stage; the path profit is the sum of the time window profit of the observation time window and the time window profit of the parent observation time window.

[0116] In the first observation phase, all nodes are connected to the starting point, and each node and the starting point share only one directed arc. Therefore, the parent observation time window of each node in the first observation phase can be set as the starting point. In this embodiment of the invention, the parent observation time window refers to the previous observation time window of the target observation time window under the path with the highest profit among all paths from the starting point to the target observation time window.

[0117] The path profit for each node can also be obtained. In this embodiment of the invention, the path profit refers to the sum of the time window profit of the target observation time window and the path profit of its parent observation time window. In this embodiment of the invention, the path profit of the starting point is set to 0.

[0118] Therefore, based on the path profit at the starting point and the time window profit of each observation time window in the first observation stage, the path profit of each observation time window in the first observation stage can be obtained.

[0119] S21243. Obtain the node with the largest path profit among the starting nodes of all directed arcs corresponding to each observation time window in the second observation stage, and use it as the parent observation time window for each observation time window; obtain the path profit for each observation time window in the second observation stage.

[0120] In the second observation phase, nodes are connected to nodes from the first observation phase, forming directed arcs. Therefore, each node may have multiple directed arcs with preceding nodes. For example... Figure 2 Node v4 is a middle node, and there are two directed arcs between it and the previous nodes: v1-v4 and v8-v4. Therefore, there are two paths from the starting point to node v4. Thus, we need to select the path with the highest profit.

[0121] Specifically, for each node in the second observation phase, we can sequentially traverse each node and all directed arcs existing in the previous observation phase. Then, we identify the target starting node with the largest path profit among the starting nodes of these directed arcs and determine it as the parent observation time window for each node. For example, for node v4, there are two directed arcs: v1-v4 and v8-v4. We can identify the starting nodes v1 and v8 of these two directed arcs with the largest path profit. If v1 has the largest path profit, then v1 is the parent observation time window for v4.

[0122] After determining the parent observation time window for all nodes in the second observation phase, the path profit for all nodes can be obtained.

[0123] S21244. Based on the path profit of each observation time window in the second observation stage, obtain the parent observation time window and path profit of each observation time window in the next observation stage, until the parent observation time window and path profit of each observation time window in the last observation stage are obtained.

[0124] Specifically, following the steps described in S21243, we can continue to obtain the parent observation time window and path profit of all nodes in the third observation stage, and then, based on the path profit of all nodes in the third observation stage, continue to obtain the parent observation time window and path profit of all nodes in the next observation stage. This continues until the parent observation time window and path profit of all nodes in the last stage are obtained.

[0125] S21245. The observation time window with the highest path profit in the last observation stage is determined as the target observation time window. All parent observation time windows corresponding to the target observation time window are obtained. Based on the target observation time window, all parent observation time windows and the endpoint, the parent observation time window path of the target observation time window is determined.

[0126] Specifically, after obtaining the node with the highest path profit in the last observation stage, we can obtain the parent observation time window (parent node) of that node, then obtain the parent nodes before that parent node, and so on, to obtain all the parent observation time windows corresponding to the target observation time window.

[0127] Based on the target observation time window, all its corresponding parent observation time windows, and the endpoint, a path can be determined, which is referred to as the parent observation time window path in this embodiment of the invention.

[0128] S21246. Determine all the first observation time windows in the parent observation time window path as the target observation time window set for the first task to be observed.

[0129] In the parent observation time window path, there may be virtual observation time windows, start points and end points, but these nodes are not real observation time windows. Therefore, it is necessary to filter out the real observation time windows in the parent observation time window path, that is, to filter out all the first observation time windows, and use these first observation time windows as the target observation time window set of the first task to be observed, that is, the planning scheme of the first task to be observed.

[0130] S2125. Continue to obtain the target observation time window set for the next task to be observed, until the target observation time window set for all tasks to be observed is obtained.

[0131] Based on step S2124 above, the steps for obtaining a set of target observation time windows for a task to be observed can be understood. Therefore, by referring to this step, the set of target observation time windows for all tasks to be observed can be obtained. Specifically, it includes the following steps:

[0132] S21251. Obtain the incompatible observation time windows of all observation time windows in the target observation time window set of the first observation task, wherein the incompatible observation time windows are observation time windows that cannot be executed simultaneously.

[0133] Let a, b∈W j Let a and b represent two different observation time windows for the j-th satellite. If the transition time between a and b does not meet the requirements for the power-on / off transition of the satellite's sensors, then they are an incompatible pair of time windows, denoted by (a, b).

[0134] Two incompatible observation windows are two observation windows that the satellite cannot execute normally, meaning they cannot perform the observation task together.

[0135] S21252. Delete the target observation time window set of the first observation task and the incompatible observation time window from the observation time window set to update the observation time window set.

[0136] Specifically, considering that the planning scheme for the first task to be observed has been completed, its corresponding observation time window can no longer be matched by subsequent tasks. Therefore, the target observation time window set of the first task to be observed can be deleted from the total observation time window set.

[0137] Incompatible observation time windows cannot jointly perform the observation task, so they also need to be removed from the total set of observation time windows.

[0138] S21253. Based on the updated set of observation time windows, obtain the target set of observation time windows for the next task to be observed.

[0139] At this point, the total set of observation time windows has been updated. Following step S2124, the next task to be observed can be designated as the first task to be observed, thereby obtaining the target observation time window set for the next task to be observed. This process continues until the target observation time window sets for all tasks to be observed are obtained.

[0140] S213. Based on the target observation time window set of all tasks to be observed, obtain the initial task planning scheme corresponding to each combination of cost coefficients.

[0141] The set of target observation time windows for all tasks to be observed can collectively form a new set, which is the initial task planning scheme corresponding to the combination of cost coefficients.

[0142] By following the steps above, the initial planning scheme for each cost coefficient combination can be obtained sequentially.

[0143] S22. Calculate the scheme profit of the initial observation task planning scheme. The formula for calculating the scheme profit is as follows:

[0144]

[0145] in:

[0146] F represents the profit of the project;

[0147] N T K represents the number of tasks to be observed. i n represents the observation frequency of the i-th observation task; i p represents the actual number of images for the i-th observation task, which is the number of the first observation time window; i This represents the observation reward for the i-th task to be observed;

[0148] D bl This represents the initial planning scheme for the observation task; w h p represents the observation time window in the initial planning scheme for the task to be observed; i Indicates the observation time window w h The cost of using it.

[0149] In step S3, the cost coefficient set is updated based on the profit of the proposed scheme. This specifically includes the following steps:

[0150] S31. Perform a selection operation on the cost coefficient set. The selection operation is as follows: sort the initial task planning schemes corresponding to each cost coefficient combination in descending order of the scheme profit, and delete the cost coefficient combinations corresponding to a preset number of initial task planning schemes that are ranked lower from the cost coefficient set.

[0151] To obtain the most profitable planning scheme, a selection operation is performed on all cost coefficient combinations to update them. This involves deleting the cost coefficient combinations corresponding to the initial observed task planning schemes with lower profits, and removing the less profitable N... C N coefficient combinations are deleted from the coefficient set. C <N B .

[0152] S32. Perform a cross operation on the cost coefficient set. The number of cross operations is a preset first threshold. The cross operation is as follows: randomly copy two cost coefficient combinations, swap the cost coefficient of any one of the two cost coefficient combinations for the task to be observed, and add the two new cost coefficient combinations to the cost coefficient set.

[0153] Wherein, the first threshold N1 can be a value less than Integers.

[0154] S33. Perform a mutation operation on the cost coefficient set. The number of mutation operations is a preset second threshold. The mutation operation is as follows: randomly copy a cost coefficient combination, update the cost coefficient of any one of the observed tasks in the cost coefficient combination to any value between 0 and 1, and add the new cost coefficient combination to the cost coefficient set.

[0155] Wherein, the second threshold N2 can be N C -2*N1.

[0156] It should be noted that the purpose of crossover and mutation is to obtain new cost coefficient combinations to replace those that were deleted. Therefore, the sum of the cost coefficient combinations obtained from the crossover operation and the new cost coefficient combinations obtained from the mutation operation should be the same as the number of cost coefficient combinations deleted, i.e., the sum should be N. C Based on this constraint, specific values ​​for the first and second thresholds can be set.

[0157] In step S4, a planning scheme for the task to be observed is obtained based on the updated set of cost coefficients. This specifically includes the following steps:

[0158] S41. Set the number of iterations. The purpose of iteration is to continuously update the combination of cost coefficients in order to obtain the planning scheme with the maximum profit.

[0159] S42. Based on the updated cost coefficient set, re-execute the step of obtaining the initial observation task planning scheme corresponding to each cost coefficient combination in the cost coefficient set. Iterate and update until the number of iterations is reached.

[0160] Specifically, for the updated cost coefficient combinations, we can continue to obtain the initial task planning scheme corresponding to each cost coefficient combination. We can jump to the aforementioned step S2122 and execute it. After obtaining the initial task planning schemes for all cost coefficient combinations, we can continue to update the cost coefficient combinations and perform the next iteration until the iteration update reaches the required number of iterations.

[0161] S43. Traverse all initial observation task planning schemes and select the one with the highest profit as the observation task planning scheme.

[0162] For all the initial observation task planning schemes obtained, the one with the highest profit can be selected as the final observation task planning scheme to be executed.

[0163] This invention also provides a multi-star, multi-stage observation mission planning system based on genetic iterative search, the system comprising:

[0164] The acquisition module is configured to acquire a set of tasks to be observed and a set of observation time windows for satellites; and to construct a set of cost coefficients, which includes several cost coefficient combinations, each of which includes the cost coefficient of the task to be observed.

[0165] The profit calculation module is configured to obtain the initial planning scheme for each combination of cost coefficients based on the set of tasks to be observed and the set of observation time windows, and to calculate the scheme profit of the initial planning scheme for the tasks to be observed.

[0166] The update module is configured to update the cost coefficient combination based on the profit of the proposed scheme;

[0167] The optimal strategy acquisition module is configured to obtain the planning scheme for the task to be observed based on the updated combination of cost coefficients.

[0168] It is understood that the planning system and planning method provided in the embodiments of the present invention correspond to the planning method described above. The explanations, examples, and beneficial effects of the relevant content can be referred to the corresponding content in the multi-star multi-stage observation mission planning method based on genetic iterative search, and will not be repeated here.

[0169] This invention also provides a computer-readable storage medium storing at least one instruction, which is loaded and executed by a processor to implement the multi-star, multi-stage observation mission planning method based on genetic iterative search as described above.

[0170] This invention also provides an electronic device, which includes:

[0171] One or more processors;

[0172] Memory; and

[0173] One or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, the programs including a multi-star, multi-stage observation mission planning method based on genetic iterative search as described above.

[0174] In summary, compared with existing technologies, it has the following beneficial effects:

[0175] This invention provides an embodiment that acquires a set of tasks to be observed and a set of satellite observation time windows, and constructs a set of cost coefficients. The cost coefficient set includes several cost coefficient combinations, each containing the cost coefficients for all tasks to be observed. Based on the set of tasks and observation time windows, an initial task planning scheme corresponding to each cost coefficient combination in the cost coefficient set is obtained, and the scheme profit of the initial task planning scheme is calculated. The cost coefficient set is updated based on the scheme profit; a task planning scheme is then obtained based on the updated cost coefficient set. Each task to be observed can have a cost coefficient set for all observation time windows. This cost coefficient allows for the adjustment of the time window cost, thereby planning for each task. Based on the profit of all planning schemes, the cost coefficients can be updated to obtain a greater profit, resulting in a final task planning scheme, thus improving the effectiveness and reliability of the planning scheme.

[0176] It should be noted that, through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms. Based on this understanding, the above technical solutions, in essence or the parts that contribute to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in various embodiments or some parts of the embodiments. Numerous specific details are set forth in the specification provided herein. However, it is understood that embodiments of the present invention can be practiced without these specific details. In some instances, well-known methods, structures, and techniques have not been shown in detail so as not to obscure the understanding of this specification.

[0177] In this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, without necessarily requiring or implying any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0178] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A multi-star, multi-stage observation mission planning method based on genetic iterative search, wherein the planning method is executed by a computer, characterized in that, Includes the following steps: Obtain the set of tasks to be observed and the set of observation time windows for satellites; construct a set of cost coefficients, which includes several cost coefficient combinations, each cost coefficient combination including the cost coefficient of the task to be observed; Based on the set of tasks to be observed and the set of observation time windows, obtain the initial task planning scheme corresponding to each cost coefficient combination in the cost coefficient set, and calculate the scheme profit of the initial task planning scheme. Update the cost coefficient set based on the profit of the proposed scheme; Obtain the planning scheme for the task to be observed based on the updated set of cost coefficients; The step of updating the cost coefficient set based on the profit of the proposed scheme includes: A selection operation is performed on the cost coefficient set. The selection operation is as follows: sort the initial task planning schemes corresponding to each cost coefficient combination according to the order of the scheme profit from large to small, and delete the cost coefficient combinations corresponding to the lower-ranked initial task planning schemes from the cost coefficient set. A crossover operation is performed on the cost coefficient set. The number of crossover operations is a preset first threshold. The crossover operation is as follows: two cost coefficient combinations are randomly copied, and the cost coefficient of any one of the observed tasks in the two cost coefficient combinations is swapped. The two new cost coefficient combinations are then added to the cost coefficient set. A mutation operation is performed on the cost coefficient set. The number of mutation operations is a preset second threshold. The mutation operation is as follows: a cost coefficient combination is randomly copied, and the cost coefficient of any one of the observed tasks in the cost coefficient combination is updated to any value between 0 and 1. The new cost coefficient combination is then added to the cost coefficient set. The process of obtaining the planning scheme for the observed task based on the updated cost coefficient combination includes: Set the number of iterations; Based on the updated set of cost coefficients, the step of obtaining the initial observation task planning scheme corresponding to each combination of cost coefficients in the set of cost coefficients is re-executed; iterative updates are performed until the number of iterations is reached. Iterate through all the initial task planning schemes and select the one with the highest profit as the task planning scheme.

2. The planning method according to claim 1, characterized in that, The initial observation task planning scheme corresponding to each cost coefficient combination in the cost coefficient set includes: Obtain the task benefits of all tasks to be observed and the usage costs of all observation time windows, and sort the tasks to be observed in descending order of the task benefits; Based on the sorting results, and using the observation time window set and the usage cost, the target observation time window set for all tasks to be observed is obtained sequentially. Based on the target observation time window set of all tasks to be observed, obtain the initial task planning scheme corresponding to each combination of cost coefficients.

3. The planning method according to claim 2, characterized in that, The step of sequentially obtaining the target observation time window set for all tasks to be observed includes: Within the set of observation time windows, obtain the first set of observation time windows corresponding to the first task to be observed; obtain the observation benefits, observation frequency, observation satellite type, and observation satellite order of the first task to be observed; Calculate the product of the usage cost of the first observation time window and the cost coefficient of the first task to be observed, and use it as the update cost of the first observation time window; Based on the first observation time window, the type of the observed satellite, and the order of the observed satellites, a directed graph of the observation time window path corresponding to the first task to be observed is obtained. The target observation time window set of the first task to be observed is obtained based on the directed graph of the observation time window path. Continue to obtain the target observation time window set for the next task to be observed, until the target observation time window set for all tasks to be observed is obtained.

4. The planning method according to claim 3, characterized in that, The process of obtaining the directed graph of the observation time window path corresponding to the first task to be observed includes: Establish a start point and an end point, and based on the observation frequency, set the observation phase for the first task to be observed; Based on the observed satellite type and the observed satellite order, the first observation time window is added to the corresponding observation stage, and a virtual observation time window is established in each observation stage; Based on preset connection conditions, the observation time windows in different observation stages are connected by directed arcs to obtain a directed graph of observation time window paths.

5. The planning method according to claim 4, characterized in that, The step of obtaining the target observation time window set for the first task to be observed based on the directed graph of the observation time window path includes: Based on the update cost, calculate the time window profit of the first observation time window, and set the time window profits of the start point, the end point and the virtual observation time window to 0; The parent observation time window of each observation time window in the first observation stage is set as the starting point, and the path profit of each observation time window in the first observation stage is obtained; the path profit is the sum of the time window profit of the observation time window and the path profit of the parent observation time window; the path profit of the starting point is 0. In the second observation phase, obtain the node with the largest path profit among the starting nodes of all directed arcs corresponding to each observation time window, and use it as the parent observation time window for each observation time window; obtain the path profit for each observation time window in the second observation phase. Based on the path profit of each observation time window in the second observation stage, obtain the parent observation time window and path profit of each observation time window in the next observation stage, until the parent observation time window and path profit of each observation time window in the last observation stage are obtained. The observation time window with the highest path profit in the last observation stage is determined as the target observation time window; all parent observation time windows corresponding to the target observation time window are obtained, and the parent observation time window path of the target observation time window is determined based on the target observation time window, all parent observation time windows and the endpoint; All first observation time windows in the parent observation time window path are determined as the target observation time window set for the first task to be observed.

6. The planning method according to claim 5, characterized in that, The process of obtaining the target observation time window set for the next task to be observed includes: Obtain the incompatible observation time windows of all observation time windows in the target observation time window set of the first observation task, wherein the incompatible observation time windows are observation time windows that cannot be executed simultaneously. The target observation time window set of the first observation task and the incompatible observation time window are deleted from the observation time window set to update the observation time window set; Based on the updated set of observation time windows, obtain the target set of observation time windows for the next task to be observed.

7. The planning method according to claim 6, characterized in that, The calculation of the scheme profit of the initial observation task planning scheme includes: in: F represents the profit of the project; Indicates the number of tasks to be observed; Indicates the first i The observation frequency of each task to be observed; Indicates the first i The actual number of imaging operations for each observation task; Indicates the first i The observational benefits of each task to be observed; This represents the initial planning scheme for the task to be observed; This represents the observation time window in the initial planning scheme for the task to be observed; Indicates the observation time window The cost of using it.

8. A multi-star, multi-stage observation mission planning system based on genetic iterative search, characterized in that, The system includes: The acquisition module is configured to acquire a set of tasks to be observed and a set of observation time windows for satellites; and to construct a set of cost coefficients, which includes several cost coefficient combinations, each of which includes the cost coefficient of the task to be observed. The profit calculation module is configured to obtain the initial planning scheme for each combination of cost coefficients in the cost coefficient set based on the set of tasks to be observed and the set of observation time windows, and to calculate the scheme profit of the initial planning scheme for tasks to be observed. The update module is configured to update the set of cost coefficients based on the profit of the proposed scheme; The optimal strategy acquisition module is configured to obtain the planning scheme for the task to be observed based on the updated set of cost coefficients. The step of updating the cost coefficient set based on the profit of the proposed scheme includes: A selection operation is performed on the cost coefficient set. The selection operation is as follows: sort the initial task planning schemes corresponding to each cost coefficient combination according to the order of the scheme profit from large to small, and delete the cost coefficient combinations corresponding to the lower-ranked initial task planning schemes from the cost coefficient set. A crossover operation is performed on the cost coefficient set. The number of crossover operations is a preset first threshold. The crossover operation is as follows: two cost coefficient combinations are randomly copied, and the cost coefficient of any one of the observed tasks in the two cost coefficient combinations is swapped. The two new cost coefficient combinations are then added to the cost coefficient set. A mutation operation is performed on the cost coefficient set. The number of mutation operations is a preset second threshold. The mutation operation is as follows: a cost coefficient combination is randomly copied, and the cost coefficient of any one of the observed tasks in the cost coefficient combination is updated to any value between 0 and 1. The new cost coefficient combination is then added to the cost coefficient set. The process of obtaining the planning scheme for the observed task based on the updated cost coefficient combination includes: Set the number of iterations; Based on the updated set of cost coefficients, the step of obtaining the initial observation task planning scheme corresponding to each combination of cost coefficients in the set of cost coefficients is re-executed; iterative updates are performed until the number of iterations is reached. Iterate through all the initial task planning schemes and select the one with the highest profit as the task planning scheme.

Images