Multi-satellite and multi-frequency earth observation mission planning method and system based on greedy search

Abstract

The present invention provides a multi-satellite, multi-frequency Earth observation mission planning method and system based on greedy search, relating to the technical field of satellite mission planning. Based on the greedy idea, the present invention first solves tasks one by one in order of mission benefits to obtain more profits. When solving each task, each task is converted into a directed graph. When assigning values to arcs in the directed graph, weights are set to non-negative values on the basis of maximizing profits, so that the maximization profit problem is converted into a problem of solving the shortest path from the start node to the end node in the graph. An existing algorithm is then used to solve a set of time windows for the task. Time windows that conflict with the planned solution are then searched for in the set of time windows. After these time windows are eliminated, the next task can be planned, ultimately obtaining an approximate optimal solution for the entire mission planning solution that is very close to the global optimal solution.

Description

Technical Field

[0001] The present invention relates to the technical field of satellite mission planning, and in particular to a multi-satellite multi-frequency earth observation mission planning method and system based on greedy search. Background Art

[0002] Earth observation satellites offer advantages such as wide coverage, long duration, freedom from time, geographic restrictions, and national boundaries, and they are not subject to human safety concerns. They play a vital role in resource exploration, environmental monitoring, military reconnaissance, and disaster relief. In practical applications, Earth observation satellites often require multi-frequency observation of targets. For example, fire-prone areas require combined observations from high- and low-orbit satellites. High-orbit satellites patrol key areas daily. Once a disaster site is identified, multiple low-orbit, high-resolution satellites are used to image the affected area multiple times. This type of mission is known as a multi-frequency observation mission.

[0003] Multi-frequency observation mission is a typical complex earth observation mission. This type of mission requires imaging the same ground point target K times. That is, given K time periods within the scheduling period H, it is required to observe the same ground point target once in each time period. At the same time, the time interval E of the imaging of two adjacent frequencies satisfies E. min ≤E≤E max , where E min 、E max The complex requirements of multi-frequency observation missions bring new challenges to satellite mission planning.

[0004] The existing multi-frequency observation tasks are inefficient and have poor scheduling effects through manual planning. Summary of the Invention

[0005] (1) Technical problems solved

[0006] In response to the shortcomings of the existing technology, the present invention provides a multi-satellite multi-frequency Earth observation mission planning method and system based on greedy search, which solves the problem of how to plan multi-satellite multi-frequency observation missions.

[0007] (2) Technical solution

[0008] To achieve the above objectives, the present invention is implemented through the following technical solutions:

[0009] In a first aspect, a multi-satellite multi-frequency Earth observation mission planning method based on greedy search comprises:

[0010] S1. Obtain task-related data; and sort the set of observation tasks T in the task-related data in non-ascending order according to the task benefits; and represents a set of multi-frequency observation tasks, t i Represents the i-th task, 1≤i≤N T , N T Indicates the number of multi-frequency observation tasks;

[0011] S2, let i = 1, the time window set D be an empty set, and the planning scheme Q be an empty set;

[0012] S3. Determine i≤N T Is it true? If so, execute S4; otherwise, execute S8;

[0013] S4, with the goal of maximizing the profit of completing all tasks, based on the i-th task t in the observed task set T i Construct a directed graph F without negative weights C ; and solve graph F C The shortest path of the image node is constructed based on the time window corresponding to the imaging node in the shortest path.

[0014] S5. Add the time windows in the time window set D to the planning scheme Q;

[0015] S6. Obtain a time window set C that conflicts with the time windows in the time window set D, and delete the time window set C and the time windows in the time window set D from the time window set W;

[0016] S7. Clear the time window set D, set i=i+1, and return to S3.

[0017] S8. Output planning scheme Q.

[0018] Furthermore, the task-related data includes:

[0019] represents the set of Earth observation satellites, s j represents the jth Earth observation satellite;

[0020] represents the time window set of all Earth observation satellites for all missions; w h Represents the hth time window, 1≤h≤N w ; Each time window includes a set of start time and end time; c h represents the time window w h Cost of use;

[0021] Represents task t i A set of time windows;

[0022] The time window set of the jth satellite;

[0023] pi represents the task benefit of completing the i-th task; and the observation task set T is arranged in non-ascending order according to the task benefit;

[0024] K i represents the observation frequency required by the i-th task;

[0025] The incompatible time window pair set O includes several incompatible time window pairs (a, b), and (a, b) represents a, b∈W j Represent two different time windows of the j-th satellite. If a and b do not meet the power-on and power-off transition time requirements of the satellite-borne sensor, they are a set of incompatible time window pairs.

[0026] E min 、E max They represent the minimum and maximum time intervals between two adjacent imaging tasks respectively.

[0027] Furthermore, the goal is to maximize the profit of completing all tasks, based on the i-th task t in the observed task set T i Construct a directed graph F without negative weights C ,include:

[0028] S4.1.1. Construction of Figure F C Nodes in , where Graph F C There is a starting node v s , an end node v e and K i frequency time intervals, each frequency time interval contains several imaging nodes and one non-imaging node;

[0029] And the construction method of imaging nodes in each frequency time interval is:

[0030] The task t i The time window set W i Each time window in is divided into a frequency time interval corresponding to a corresponding time period according to the start time as an imaging node within the frequency time interval;

[0031] S4.1.2. Construct directed arcs connecting each node, and the directed arc construction rules are:

[0032] Set the starting node v s The imaging nodes and non-imaging nodes in the first frequency time interval are connected with directed arcs, and the arrows point to the time window in the first frequency;

[0033] The end node v e Connected to the imaging node and non-imaging node in the last frequency time interval, with the arrow pointing to the end node ve ;

[0034] The time interval between the end time of the time window corresponding to the imaging node in the kth frequency time interval and the start time of the time window corresponding to the imaging node in the k+1th frequency time interval is taken as the imaging time interval E between the two imaging nodes; then determine whether the imaging time interval E between the time window corresponding to the imaging node in the kth frequency time interval and the time window corresponding to the imaging node in the k+1th frequency time interval satisfies E min ≤E≤E max If yes, connect the two imaging nodes with a directed arc, with the arrow pointing to the imaging node in the k+1th frequency time interval; otherwise, do not connect; where k = 1, 2, ..., K i -1;

[0035] All imaging nodes in each frequency time interval are connected to non-imaging nodes in the next frequency time interval. At the same time, non-imaging nodes in each frequency time interval are connected to all imaging nodes in the next frequency time interval.

[0036] S4.1.3. Constructing Figure F C The weights of each directed arc in , and the weight assignment rules are:

[0037] For non-image nodes: Let graph F C The profit corresponding to the non-image node is g s =0;

[0038] For imaging nodes: first calculate the profit corresponding to each imaging node, and the profit g of the imaging node m The calculation formula is:

[0039]

[0040] Among them, c m Represents the imaging node v m The usage cost of the corresponding time window, if the imaging node v m The corresponding time window is w h , then c m =c h ;

[0041] For the end node: Let the end node v e Profit g e =M, where M is a constant greater than the profit of all nodes;

[0042] Let graph F C Slave node v p Points to node v q The directed arc (v p ,vq ) has a weight of g pq =Mg q .

[0043] Furthermore, the above-mentioned C The weight and the minimum shortest path are used to obtain the time window set D, including:

[0044] Constructing Graph F C The objective function and constraints are:

[0045]

[0046] Among them, x pq Indicates whether there is a slave node v p Points to node v q If the directed arc is x, then pq =1, otherwise, x pq =0;

[0047] A represents the set of nodes in the graph;

[0048] Constraints include:

[0049] Figure F C The solution satisfies that the starting node has only one outflow arc:

[0050]

[0051] Among them, v n :(v s ,v n )∈C represents the graph F C All the nodes from the starting node v s outflow arc;

[0052] Figure F C The solution satisfies that the end node has only one inflow arc:

[0053]

[0054] Among them, v m :(v m ,v e )∈C represents the graph F C All inflow end nodes v e arc;

[0055] Figure F C The solution satisfies that the remaining nodes except the start node and the end node satisfy:

[0056]

[0057] Among them, vq :(v p ,v q )∈C represents the graph F C All inflow nodes v y arc;

[0058] v q :(v q ,v p )∈C represents the graph F C All slave nodes v y outflow arc;

[0059] Figure F C The solution satisfies the constant number of directed arcs K i +1.

[0060] Furthermore, obtaining a time window set C that conflicts with a time window in the time window set D, and deleting the time window set C and the time windows in the time window set D from the time window set W, includes:

[0061] S6.1. Let the conflict time window set C be an empty set, d = 1, o = 1;

[0062] S6.2. Obtain the number of time windows |D| in the time window set D and the number of incompatible pairs |O| in the incompatible time window pair set O;

[0063] S6.3. Determine whether d ≤ |D| holds. If so, go to S6.4; otherwise, go to S6.9.

[0064] S6.4. Determine whether o ≤ |O| holds. If so, go to S6.5; otherwise, go to S6.8.

[0065] S6.5. Determine whether the oth incompatible time window pair in the incompatible time window pair set O contains the dth time window in the time window set D. If so, go to S6.6; otherwise, go to S6.7.

[0066] S6.6. Add the time window that forms an incompatible time window pair with the dth time window in the time window set D to the conflicting time window set C;

[0067] S6.7. Update o = o + 1 and go to S6.4.

[0068] S6.8. Update d = d + 1, and go to S6.3;

[0069] S6.9. Output the conflict time window set C.

[0070] Furthermore, the solution graph F C The shortest path includes: solving using Dijkstra algorithm.

[0071] In the second aspect, a multi-satellite multi-frequency earth observation mission planning system based on greedy search is provided, wherein the system includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and the steps of the above method are implemented when the processor executes the computer program.

[0072] (3) Beneficial effects

[0073] The present invention provides a multi-satellite, multi-frequency Earth observation mission planning method and system based on greedy search. Compared with the existing technology, it has the following advantages:

[0074] The present invention is based on the greedy idea and first solves the tasks one by one in the order of task benefits so as to obtain more profits. When solving each task, each task is converted into a directed graph, and when assigning values to the arcs of the directed graph, on the basis of satisfying the maximization of profit, the weight is set to a non-negative value, so that the maximization profit problem is converted into a problem of solving the shortest path from the start node to the end node in the graph, and then the existing algorithm is used to solve the time window set of the task, and then the time window that conflicts with the planned plan is found in the time window set. After eliminating it, the next task can be planned, and finally an approximate optimal solution of the entire task planning scheme is obtained which is very close to the global optimal solution. BRIEF DESCRIPTION OF THE DRAWINGS

[0075] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0076] Figure 1 is a flow chart of an embodiment of the present invention;

[0077] Figure 2 FIG. F of an embodiment of the present invention C Schematic diagram of;

[0078] Figure 3 This is a flow chart of a time window conflict determination method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0079] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention are clearly and completely described. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. All other embodiments derived by persons of ordinary skill in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0080] The embodiments of the present application solve the problem of how to plan multi-satellite multi-frequency observation missions by providing a multi-satellite multi-frequency Earth observation mission planning method and system based on greedy search.

[0081] The technical solution in the embodiments of the present application is to solve the above technical problems, and the overall idea is as follows:

[0082] For a task set consisting of a large number of multi-frequency observation tasks, different tasks will compete for limited satellite resources. The tasks planned first will occupy some satellite resources, which in turn affects the available time window of the tasks planned later. Therefore, the order of task planning has a very significant impact on the final planning scheme. To this end, based on the greedy idea, the present invention first sorts the tasks in the task set according to their benefits, and gives priority to planning tasks with greater benefits so that the final planning scheme can obtain more profits.

[0083] Based on a greedy approach to sorting all tasks in a task set, this paper designs a graph-based conversion method, converting each task into a directed graph. The profit of each time window is calculated based on the task's revenue, frequency, and time window usage cost, and arcs are weighted according to the time window profit. To facilitate the solution, when assigning weights to the graph, the present invention converts the equivalent solution into a non-negatively weighted shortest path problem. This allows the Dijkstra method from graph theory to solve the shortest path from the starting node to the end node in the graph.

[0084] After each task planning is completed, in order to screen the available time windows for subsequent tasks, the present invention also designs a heuristic conflict judgment method to search for time windows that conflict with the planned scheme in the time window set. After eliminating the conflicting time windows, the available time window set for subsequent tasks can be determined, and then the next task can be planned, and finally the planning scheme for all tasks can be obtained.

[0085] In order to better understand the above technical solution, the above technical solution will be described in detail below with reference to the accompanying drawings and specific implementation methods.

[0086] Example 1:

[0087] like Figure 1 As shown, the present invention provides a multi-satellite multi-frequency Earth observation mission planning method based on greedy search, the method comprising:

[0088] S1. Obtain task-related data; and sort the set of observation tasks T in the task-related data in non-ascending order according to the task benefits; and represents a set of multi-frequency observation tasks, t i Represents the i-th task, 1≤i≤NT , N T Indicates the number of multi-frequency observation tasks;

[0089] S2, let i = 1, the time window set D be an empty set, and the planning scheme Q be an empty set;

[0090] S3. Determine i≤N T Is it true? If so, execute S4; otherwise, execute S8;

[0091] S4, with the goal of maximizing the profit of completing all tasks, based on the i-th task t in the observed task set T i Construct a directed graph F without negative weights C ; and solve graph F C The shortest path of the image node is constructed based on the time window corresponding to the imaging node in the shortest path.

[0092] S5. Add the time windows in the time window set D to the planning scheme Q;

[0093] S6. Obtain a time window set C that conflicts with the time windows in the time window set D, and delete the time window set C and the time windows in the time window set D from the time window set W;

[0094] S7. Clear the time window set D, set i=i+1, and return to S3.

[0095] S8. Output planning scheme Q.

[0096] The beneficial effects of this embodiment are:

[0097] The present invention is based on the greedy idea and first solves the tasks one by one in the order of task benefits so as to obtain more profits. When solving each task, each task is converted into a directed graph, and when assigning values to the arcs of the directed graph, on the basis of satisfying the maximization of profit, the weight is set to a non-negative value, so that the maximization profit problem is converted into a problem of solving the shortest path from the start node to the end node in the graph, and then the existing algorithm is used to solve the time window set of the task, and then the time window that conflicts with the planned plan is found in the time window set. After eliminating it, the next task can be planned, and finally an approximate optimal solution of the entire task planning scheme is obtained which is very close to the global optimal solution.

[0098] The implementation process of the embodiment of the present invention is described in detail below:

[0099] definition:

[0100] Multi-frequency observation task: The scheduling period H is divided into K frequency time intervals according to the observation frequency K. In each frequency time interval, the same ground point target must be observed once. The imaging time interval E of two adjacent observation frequencies satisfies E min ≤E≤E max , where E min 、E max Respectively represent the minimum and maximum time intervals between two adjacent imaging tasks. Therefore, the method includes the following steps:

[0101] S1. Obtain task-related data.

[0102] Specifically, the task-related data required includes:

[0103] represents a set of multi-frequency observation tasks, t i Represents the i-th task, 1≤i≤N T , N T Indicates the number of multi-frequency observation tasks;

[0104] represents the set of Earth observation satellites, s j represents the jth Earth observation satellite, each of which carries an imaging sensor;

[0105] represents the time window set of all Earth observation satellites for all missions; w h Represents the hth time window, 1≤h≤N w ; Each time window includes a start time and an end time; c h represents the time window w h In this embodiment, the cost may be physical quantities such as the satellite's power consumption and storage space;

[0106] Represents task t i A set of time windows;

[0107] The time window set of the jth satellite;

[0108] p i represents the task benefit of completing the i-th task; and the observation task set T is arranged in non-ascending order according to the task benefit;

[0109] K i represents the observation frequency required by the i-th task;

[0110] The incompatible time window pair set O includes several incompatible time window pairs (a, b), and (a, b) represents a, b∈Wj They represent two different time windows of the j-th satellite. If a and b do not meet the power-on and power-off transition time requirements of the satellite-borne sensor, they are a set of incompatible time window pairs.

[0111] E min 、E max They represent the minimum and maximum time intervals between two adjacent imaging tasks respectively.

[0112] S2. Initialize algorithm parameters.

[0113] Specifically, let i=1, the time window set D be an empty set, and the planning scheme Q be an empty set.

[0114] Among them, the time window set D represents the i-th task t in the observation task set T. i The time window obtained after planning can be used to determine the task t through the time window in the time window set D. i By which satellite and at what time.

[0115] The planning scheme Q represents the time window corresponding to the planned tasks.

[0116] S3. Determine i≤N T Is it true? If so, it means that there are still tasks in the observation task set T that have not been planned, and S8 is executed to plan the next task; otherwise, it means that all tasks in the observation task set T have been planned, and S4 is executed.

[0117] S4, with the goal of maximizing the profit of completing all tasks, based on the i-th task t in the observed task set T i Construct a directed graph F without negative weights C ; and solve graph F C The shortest path is constructed by constructing a time window set D based on the time windows corresponding to the imaging nodes in the shortest path.

[0118] Specifically, the directed graph F C The construction mainly includes: constructing graph F C Nodes in, construct directed arcs connecting each node and construct graph F C The weights of each directed arc in .

[0119] S4.1.1. Constructing Figure F C Nodes in .

[0120] Among them, Figure F C There is a starting node v s , an end node v e and K ifrequency time intervals, each frequency time interval contains several imaging nodes and one non-imaging node;

[0121] And the construction method of imaging nodes in each frequency time interval is:

[0122] The task t i The time window set W i Each time window in is divided into a frequency time interval corresponding to a corresponding time period according to the start time as an imaging node within the frequency time interval.

[0123] Obviously, each imaging node corresponds to a time window within the frequency time interval, and the time window represents an optional observation time of the frequency corresponding to the task.

[0124] S4.1.2. Construct directed arcs connecting each node.

[0125] And the directed arc construction rules are:

[0126] Set the starting node v s The imaging nodes and non-imaging nodes in the first frequency time interval are connected with directed arcs, and the arrows point to the time window in the first frequency;

[0127] The end node v e Connected to the imaging node and non-imaging node in the last frequency time interval, with the arrow pointing to the end node v e ;

[0128] The time interval between the end time of the time window corresponding to the imaging node in the kth frequency time interval and the start time of the time window corresponding to the imaging node in the k+1th frequency time interval is taken as the imaging time interval E between the two imaging nodes; then determine whether the imaging time interval E between the time window corresponding to the imaging node in the kth frequency time interval and the time window corresponding to the imaging node in the k+1th frequency time interval satisfies E min ≤E≤E max If yes, connect the two imaging nodes with a directed arc, with the arrow pointing to the imaging node in the k+1th frequency time interval; otherwise, do not connect; where k = 1, 2, ..., K i -1;

[0129] All imaging nodes in each frequency time interval are connected to non-imaging nodes in the next frequency time interval. At the same time, non-imaging nodes in each frequency time interval are connected to all imaging nodes in the next frequency time interval.

[0130] After constructing the directed arcs, it is necessary to assign weights to the arcs. Since the goal is to maximize the profit of completing all tasks, it is obvious that in this embodiment, the specific meaning of profit can be defined as needed. For example, it can be physical quantities such as the satellite's power consumption and storage space. The objective function can be expressed as follows:

[0131]

[0132] Among them, x pq Indicates whether there is a slave node v p Points to node v q If the directed arc is x, then pq =1, otherwise, x pq =0;

[0133] g pq Represents a directed arc (v p ,v q )’s weight;

[0134] A represents the set of nodes in the graph.

[0135] Obviously, under this objective function, if profit is directly used as the weight of the arc, it cannot be solved using the existing algorithm. In order to facilitate subsequent solution calculations, the profit maximization problem needs to be converted into a non-negative weight shortest path problem, and the weights need to be further adjusted.

[0136] Assuming that profit is directly used as the weight of the arc, a directed graph F can be constructed A , where F A With F C The nodes and directed arcs are the same, and the constraints corresponding to the objective function with profit maximization as the goal are:

[0137] Figure F A The solution satisfies that the starting node has only one outflow arc:

[0138]

[0139] Among them, v n :(v s ,v n )∈A represents graph F A All the nodes from the starting node v s outflow arc;

[0140] Figure F A The solution satisfies that the end node has only one inflow arc:

[0141]

[0142] Among them, v m :(vm ,v e )∈A represents graph F A All inflow end nodes v e arc;

[0143] Figure F A The solution satisfies that all nodes except the start and end nodes satisfy (that is, the number of inflow arcs and outflow arcs is equal, both 0 or 1):

[0144]

[0145] Among them, v q :(v p ,v q )∈A represents graph F A All inflow nodes v y Arc v q :(v q ,v p )∈A represents graph F A All slave nodes v y outflow arc;

[0146] Figure F A The solution satisfies the constant number of directed arcs K i +1.

[0147] And the weight assignment rule is:

[0148] Let graph F A The profit corresponding to the non-image node is g s =0;

[0149] Figure F A The profit of each imaging node can be expressed as:

[0150]

[0151] Among them, g m Represents the imaging node v m Profit, c m Represents the imaging node v m The usage cost of the corresponding time window.

[0152] Let the end node v e Profit g e =M, where M is a constant greater than the profit of all nodes;

[0153] Then for the directed graph F A , directed arc (v p ,v q ) is denoted as g pq =g q .

[0154] Then for Figure F A The weights of all directed arcs in are transformed for the first time, and the transformation formula is as follows:

[0155] g pq =-g pq

[0156] Based on the weights after the first transformation, we get the new graph F B Figure F B With Figure F A Compared to , only the weights of the directed arcs are different. B The goal is to minimize the weight of the path, Figure F B The solution also satisfies Figure F A Constraints, so Figure F B With Figure F A Same solution.

[0157] Finally, based on Figure F B The weights of all directed arcs in the graph are transformed for the second time to obtain graph F. C The weights of all directed arcs in , and the formula for the secondary transformation is as follows:

[0158] g pq =M+g pq

[0159] That is F C The directed arc (v p ,v q ) can be recorded as: g pq =Mg q , g q Represents the profit corresponding to node q.

[0160] Figure F B With Figure F C Compared to Figure F, only the weights of the directional arcs are different. C The goal is to minimize the total path length, and Figure F C The weights of all arcs in Figure F are not negative. C The solution also satisfies Figure F A Constraints, so Figure F C With Figure F B Same solution, so Figure F C With Figure F A Same solution. So far, the original profit maximization problem has been transformed into the shortest path problem.

[0161] Figure F C The objective function is:

[0162]

[0163] And the constraints are also:

[0164] Figure F C The solution satisfies that the starting node has only one outflow arc:

[0165]

[0166] Among them, v n :(v s ,v n )∈C represents the graph F C All the nodes from the starting node v s outflow arc;

[0167] Figure F C The solution satisfies that the end node has only one inflow arc:

[0168]

[0169] Among them, v m :(v m ,v e )∈C represents the graph F C All inflow end nodes v e arc;

[0170] Figure F C The solution satisfies that all nodes except the start and end nodes satisfy (that is, the number of inflow arcs and outflow arcs is equal, both 0 or 1):

[0171]

[0172] Among them, v q :(v p ,v q )∈C represents the graph F C All inflow nodes v y Arc v q :(v q ,v p )∈C represents the graph F C All slave nodes v y outflow arc;

[0173] Figure F C The solution satisfies the constant number of directed arcs K i +1.

[0174] Based on the above arguments, S4.1.3, construct Figure F C The weights of each directed arc in , and the weight assignment rules are:

[0175] For non-image nodes: Let graph F C The profit corresponding to the non-image node is g s=0;

[0176] For imaging nodes: first calculate the profit corresponding to each imaging node, and the profit g of the imaging node m The calculation formula is:

[0177]

[0178] Among them, c m Represents the imaging node v m The usage cost of the corresponding time window, if the imaging node v m The corresponding time window is w h , then c m =c h ;

[0179] For the end node: Let the end node v e Profit g e =M, where M is a constant greater than the profit of all nodes;

[0180] Let graph F C Slave node v p Points to node v q The directed arc (v p ,v q ) has a weight of g pq =Mg q .

[0181] At this point, Figure F is completed. C For example, the scheduling period H can be from 0:00 to 24:00 a day. Assuming that the observation frequency of a task is K=3, it is divided into three frequency time intervals: 0:00 to 8:00, 8:00 to 17:00, and 17:00 to 24:00. The time interval between two adjacent observation actions is between 5-8 hours. Then the constructed graph F C like Figure 2 As shown in the figure, the start and end nodes are represented by black triangles. Solid dots represent imaging nodes, and hollow dots represent non-imaging nodes. Unidirectional arrows between nodes represent directed arcs. For example, the time interval between imaging opportunities v2 and v3 is less than 5 hours, so there is no arc connecting them. The first frequency interval contains two imaging nodes and one non-imaging node, indicating that after searching, the only two selectable time windows within this time window are v1 and v2.

[0182] After constructing Figure F C After that, the time windows corresponding to the imaging nodes in the shortest path can be added to the time window set D.

[0183] And solve graph F C The shortest path is constructed by constructing a time window set D based on the time windows corresponding to the imaging nodes in the shortest path.

[0184] Specifically, Dijkstra algorithm can be used to solve the graph F C The shortest path from the starting node to the ending node.

[0185] S5. Add the time windows in the time window set D to the planning scheme Q.

[0186] S6. Find a time window set C that conflicts with the time windows in the time window set D, and delete the time window set C and the time windows in the time window set D from the time window set W.

[0187] Specifically, you can use Figure 3 The conflict judgment method shown is called to find a conflicting time window set C that conflicts with a time window in a time window set D, and the steps include:

[0188] S6.1. Let the conflict time window set C be an empty set, d = 1, o = 1;

[0189] S6.2. Obtain the number of time windows |D| in the time window set D and the number of incompatible pairs |O| in the incompatible time window pair set O;

[0190] S6.3. Determine whether d ≤ |D| holds. If so, go to S6.4; otherwise, go to S6.9.

[0191] S6.4. Determine whether o ≤ |O| holds. If so, go to S6.5; otherwise, go to S6.8.

[0192] S6.5. Determine whether the oth incompatible time window pair in the incompatible time window pair set O contains the dth time window in the time window set D. If so, go to S6.6; otherwise, go to S6.7.

[0193] S6.6. Add the time window that forms an incompatible time window pair with the dth time window in the time window set D to the conflicting time window set C;

[0194] S6.7. Update o = o + 1 and go to S6.4.

[0195] S6.8. Update d = d + 1, and go to S6.3;

[0196] S6.9. Output the conflict time window set C.

[0197] S7. Clear the time window set D, set i=i+1, and return to S3.

[0198] S8. Output planning scheme Q.

[0199] Example 2

[0200] The present invention also provides a multi-satellite multi-frequency Earth observation mission planning system based on greedy search, the system comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the following steps when executing the computer program:

[0201] S1. Obtain task-related data; and sort the set of observation tasks T in the task-related data in non-ascending order according to the task benefits; and represents a set of multi-frequency observation tasks, t i Represents the i-th task, 1≤i≤N T , N T Indicates the number of multi-frequency observation tasks;

[0202] S2, let i = 1, the time window set D be an empty set, and the planning scheme Q be an empty set;

[0203] S3. Determine i≤N T Is it true? If so, execute S4; otherwise, execute S8;

[0204] S4, with the goal of maximizing the profit of completing all tasks, based on the i-th task t in the observed task set T i Construct a directed graph F without negative weights C ; and solve graph F C The shortest path of , and the time window set D is obtained;

[0205] S5. Add the time windows in the time window set D to the planning scheme Q;

[0206] S6. Obtain a time window set C that conflicts with the time windows in the time window set D, and delete the time window set C and the time windows in the time window set D from the time window set W;

[0207] S7. Clear the time window set D, set i=i+1, and return to S3.

[0208] S8. Output planning scheme Q.

[0209] It can be understood that the multi-satellite multi-frequency earth observation mission planning system based on greedy search provided in an embodiment of the present invention corresponds to the multi-satellite multi-frequency earth observation mission planning method based on greedy search mentioned above. The explanation, examples, beneficial effects, etc. of the relevant contents can refer to the corresponding contents in the multi-satellite multi-frequency earth observation mission planning method based on greedy search, and will not be repeated here.

[0210] In summary, compared with the prior art, the present invention has the following beneficial effects:

[0211] ① Based on the greedy idea, the present invention first solves the tasks one by one in the order of task benefits in order to obtain more profits, and when solving each task, each task is converted into a directed graph, and when assigning values to the arcs of the directed graph, on the basis of satisfying the maximization of profits, the weights are set to non-negative values, so that the maximization of profit problem is converted into the problem of solving the shortest path from the start node to the end node in the graph, and then the existing algorithm is used to solve the time window set of the task, and then the time window that conflicts with the planned plan is found in the time window set, and after eliminating it, the next task can be planned, and finally an approximate optimal solution of the entire task planning scheme is obtained which is very close to the global optimal solution.

[0212] It should be noted that, 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 the necessary general hardware platform. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which can be stored in a storage medium such as ROM / RAM, a disk, an optical disk, etc., and includes a number of instructions for enabling 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 certain parts of the embodiment. In this article, relational terms such as first and second are merely used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "comprise", "include" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. Without further constraints, an element defined by the phrase "comprises a..." does not preclude the existence of additional identical elements in the process, method, article or apparatus that includes the element.

[0213] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A multi-satellite and multi-frequency Earth observation mission planning method based on greedy search, characterized in that: The method includes: S1. Obtain task-related data; and perform observation task set on task-related data. Sort by task payoff in non-ascending order; and , represents a set of multi-frequency observation tasks, Indicates the tasks, , Indicates the number of multi-frequency observation tasks; S2, order i =1, time window set is an empty set, planning scheme is an empty set; S3. Judgment Is it true? If so, execute S4; otherwise, execute S8; S4, with the goal of maximizing the profit of completing all tasks, based on the observation task set The tasks Constructing a directed graph without negative weights ; and solve the graph The shortest path is constructed based on the time windows corresponding to the imaging nodes in the shortest path. ; S5. Aggregate the time windows Add the time window in the planning scheme middle; S6. Acquisition and time window collection The set of time windows with conflicting time windows in , and the time window set With time window set The time window in the time window set Delete in; S7. Clear the time window set , and order , then return to S3; S8. Output planning scheme ; The goal is to maximize the profit of completing all tasks, based on the observation task set The tasks Constructing a directed graph without negative weights ,include: S4.1.

1. Construction of the graph The nodes in the graph Includes a starting node , an end node and frequency time intervals, each frequency time interval contains several imaging nodes and one non-imaging node; S4.1.

2. Construct directed arcs connecting each node, and the directed arc construction rules are: The starting node The imaging nodes and non-imaging nodes in the first frequency time interval are connected with directed arcs, and the arrows point to the time window in the first frequency; The end node Connected to the imaging node and non-imaging node in the last frequency time interval, with the arrow pointing to the end node ; The first k The end time of the time window corresponding to the imaging node in the frequency time interval is the same as the end time of the k The time interval between the start times of the time windows corresponding to the imaging nodes in the +1 frequency time interval is used as the imaging time interval between the two imaging nodes. ; Then judge the k The time window corresponding to the imaging node in the frequency time interval is k +1 imaging time interval of the time window corresponding to the imaging node in the frequency time interval Is it satisfied If so, connect the two imaging nodes with a directed arc, with the arrow pointing to the k +1 imaging node in the frequency time interval; otherwise, no connection; where, ; S4.1.

3. Build a graph The weights of each directed arc in , and the weight assignment rules are: For non-image nodes: Let the graph The profit corresponding to the non-image node is ; For imaging nodes: first calculate the profit corresponding to each imaging node, and the profit of the imaging node The calculation formula is: in, Represents an imaging node The usage cost of the corresponding time window, if the imaging node The corresponding time window is ,but ; For the end node: Let the end node Profit ,in is a constant greater than the profit of all nodes; Order diagram Middle slave node Pointing to a node Directed arcs The weight is , Representation node q The corresponding profit.

2. The method for planning multi-satellite and multi-frequency Earth observation missions based on greedy search according to claim 1, wherein: The task-related data includes: , represents the set of Earth observation satellites, Indicates the Earth observation satellites; Represents the set of time windows for all Earth observation satellites for all missions; Indicates the time windows, ;Each time window includes a set of start time and end time; Represents a time window Cost of use; Indicates a task A set of time windows; No. A set of time windows for satellites; Indicates completion of The task benefits of each task; and the observation task set Arrange in non-ascending order according to task benefits; Indicates the The observation frequency required by the task; Incompatible time window pair set , the set of incompatible time window pairs There are several incompatible time window pairs ,and Indication command Respectively represent If there are two different time windows for each satellite If the time window between them does not meet the power-on / off transition time requirement of the satellite-borne sensor, then it is a set of incompatible time windows; 、 They represent the minimum and maximum time intervals between two adjacent imaging tasks respectively.

3. The method for planning multi-satellite and multi-frequency Earth observation missions based on greedy search according to claim 2, wherein: The construction method of imaging nodes in each frequency time interval is: The task Time window set Each time window in is divided into a frequency time interval corresponding to a corresponding time period according to the start time as an imaging node within the frequency time interval; All imaging nodes in each frequency time interval are connected to non-imaging nodes in the next frequency time interval. At the same time, non-imaging nodes in each frequency time interval are connected to all imaging nodes in the next frequency time interval.

4. The method for planning multi-satellite and multi-frequency Earth observation missions based on greedy search according to claim 2, wherein: The graph-based The weight and the minimum shortest path are used to obtain the time window set ,include: Build a graph The objective function and constraints are: in, Indicates whether a slave node exists Pointing to a node If the directed arc is ,otherwise, ; Representation node q corresponding profits; C Representation diagram The node set in ; Constraints include: picture The solution satisfies that the starting node has only one outflow arc: in, Representation diagram All from the starting node outflow arc; picture The solution satisfies that the end node has only one inflow arc: in, Representation diagram All inflow end nodes arc; picture The solution satisfies that the remaining nodes except the start node and the end node satisfy: Among them, the number of inflow arcs and outflow arcs is equal, both are 0 or 1, Representation diagram All inflow nodes arc; Representation diagram All slave nodes outflow arc; picture The solution satisfies the constant number of directed arcs .

5. The method for planning multi-satellite and multi-frequency Earth observation missions based on greedy search according to claim 2, wherein: The acquisition and time window set The set of time windows with conflicting time windows in , and the time window set With time window set The time window in the time window set Deleted, including: S6.

1. Let the conflict time window set is an empty set, , ; S6.

2. Get time window set The number of time windows , the set of incompatible time window pairs The number of incompatible pairs in ; S6.

3. Judgment Is it true? If so, go to S6.4; otherwise, go to S6.

9. S6.

4. Judgment Is it true? If so, go to S6.5; otherwise, go to S6.

8. S6.

5. Determine the set of incompatible time windows The Whether the incompatible time window pairs contain the time window set The If yes, go to S6.6, otherwise go to S6.7; S6.6, will be combined with the time window The The time windows that form an incompatible time window pair are added to the conflict time window set. ; S6.7, Update , go to S6.4; S6.8, Update , go to S6.3; S6.

9. Output conflict time window set .

6. The method for planning multi-satellite and multi-frequency Earth observation missions based on greedy search according to claim 1, wherein: The solution diagram The shortest path includes: solving using Dijkstra algorithm.

7. A multi-satellite multi-frequency Earth observation mission planning system based on greedy search, the system 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 computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

Images