Multi-satellite joint planning method and system considering task splitting

Abstract

The multi-satellite joint planning method and system considering task splitting provided by the present invention construct a task planning model. The model covers the planning of both splittable observation tasks and inseparable observation tasks, and designs phased algorithms for inseparable and splittable task planning respectively, taking into full account various constraints such as satellite resource capability level and task timeliness requirements. A large number of numerical experiments have proved that by splitting the tasks, the possibility of collaboration between multiple satellites is provided, the task completion rate is improved, the benefits of multi-satellite joint observation are significantly improved, and ultimately the observation efficiency of satellites is greatly improved.

Description

Technical Field

[0001] The present invention belongs to the field of satellite mission planning, and in particular relates to a multi-satellite joint planning method and system considering mission splitting. Background Art

[0002] As an important space-based platform, artificial earth satellites use their own payloads (visible light, infrared, synthetic aperture radar, etc.) to complete earth observation tasks such as target observation, environmental monitoring, and resource census. In the environment where the number of satellites in operation is rapidly increasing, satellite constellations are constantly being established, and the collaborative work of multiple satellites is becoming a trend, the satellite observation mode has also evolved from the previous simple single observation and simple collaborative observation of multiple satellites to a more complex multi-satellite combination collaborative observation mode. The multi-satellite combination collaboration changes the previous situation of relative independence and lack of mutual coordination between satellites through the close cooperation of multiple satellites in task processing and execution, enabling satellites to have the ability to complete more complex observation tasks.

[0003] A single space-based platform can play an important role in many fields such as agriculture and industry. However, since its mission execution is limited by resource capabilities and mission execution requirements, the use of multiple space-based platforms to work together can greatly improve the mission completion capability and efficiency. Satellites are limited by orbital and payload conditions during their in-orbit operation. Except for geosynchronous orbit satellites, the visible time window of other satellites for observing targets on the earth is between tens of seconds and minutes. The length of the visible time window may be shorter than the duration required by the mission. If collaborative planning is not adopted, large-scale, long-duration missions will not be successfully executed and can only be abandoned. Summary of the invention

[0004] The object of the present invention is to provide a multi-satellite joint planning method and system considering task splitting, so as to solve at least one of the above-mentioned technical problems existing in the prior art.

[0005] In order to solve the above technical problems, the present invention provides a multi-satellite joint planning method considering task splitting, comprising the following steps:

[0006] Step 1: Obtain the task set to be planned and the satellite set participating in the task execution;

[0007] Step 2: Divide the task set in step 1 into a divisible observation task set and an indivisible observation task set according to whether the task is divisible;

[0008] Step 3: According to the task splitting method, the splittable observation task set in step 2 is divided into a split subtask set and a task set that does not need to be split;

[0009] Step 4: Based on the input of the non-divisible observed task set, the split sub-task set and the task set that does not need to be split in steps 2 to 3, the task planning model is used to solve the optimal solution to obtain the planning solution output;

[0010] Step 5: Output the result of step 4 as the overall planning solution.

[0011] Furthermore, the task planning model in step 4 includes an objective function formula:

[0012] ;

[0013] Among them, SAT represents the satellite set, i represents the i-th satellite;

[0014] Represents a set of splittable observation tasks, subtasks;

[0015] Indicates time window resources, total Visible time windows;

[0016] is the decision variable, indicating the satellite Indivisible observation tasks or tasks that do not need to be split Arranged in time window, when the task is successfully scheduled ,otherwise , can also be set to any other value;

[0017] is the decision variable, indicating the satellite Splittable observation tasks Subtasks Arranged in time window, when the subtask is successfully scheduled ,otherwise , can also be defined as any other value;

[0018] It represents the benefits obtained by the observation task that cannot be split or does not need to be split; represents the benefit obtained from the splittable observation task;

[0019] The constraints of the objective function formula are as follows:

[0020] (1) The observation task cannot be split or does not need to be split, and the actual start time is later than the earliest allowed execution time;

[0021] (2) For observation tasks that cannot be split or do not need to be split, the actual end time must be earlier than the latest allowed end time;

[0022] (3) The actual start time of the subtask of the splittable observation task is later than the earliest allowed execution time of the splittable observation task;

[0023] (4) The actual end time of the subtask of the splittable observation task must be earlier than the latest allowed execution time of the splittable observation task;

[0024] (5) The observation task cannot be split or does not need to be split, and needs to be started within the visible time window of the observation target;

[0025] (6) The observation task cannot be split or does not need to be split, and the task needs to be completed within the visible time window of the observation target;

[0026] (7) Any subtask of a splittable observation task needs to be started within the visible time window of the observation target;

[0027] (8) Any subtask of the splittable observation task needs to be completed within the visible time window of the observation target;

[0028] (9) The observation task cannot be split or does not need to be split, and the actual observation time must be equal to the required observation time;

[0029] (10) The sum of the actual observation time of each subtask of the divisible observation task must be equal to the required observation time of the divisible observation task;

[0030] (11) Any two observation tasks that cannot be split or do not need to be split must meet the minimum satellite observation interval requirements;

[0031] (12) The execution of any two subtasks of the separable observation task meets the minimum observation interval requirement of the satellite;

[0032] (13) Any observation task that cannot be split or does not need to be split, and the subtasks of the splittable observation task meet the minimum satellite observation interval requirements;

[0033] (14) Each indivisible observation task or task that does not need to be split is executed at most once;

[0034] (15) Each subtask of a separable observation task is executed at most once.

[0035] Among them, constraints (1)-(4) describe that task execution must meet the time requirements of the task itself; constraints (5)-(8) describe that task execution must ensure visibility to the target and start and end within the visible time window; constraints (9)-(10) describe that the task execution time must be consistent with the required duration; constraints (11)-(13) describe the restrictions on task scheduling due to the minimum time interval; and constraints (14)-(15) describe the restrictions on the number of task executions.

[0036] Furthermore, the method for solving the task planning model in step 4 is:

[0037] In the first stage, the inseparable observation tasks and the tasks that do not need to be split are directly planned to obtain planning schemes for the inseparable tasks and the tasks that do not need to be split, and the remaining time windows after the inseparable tasks and the tasks that do not need to be split are calculated;

[0038] In the second stage, the split subtasks are planned to obtain the planning scheme of the splittable tasks;

[0039] Finally, the planning scheme of the inseparable tasks is sent to step 5 together with the planning scheme of the divisible tasks as the overall planning scheme output.

[0040] Furthermore, the inseparable tasks and tasks that do not need to be split are planned by an algorithm, such as the improved fireworks algorithm (FWA for short, an intelligent evolutionary algorithm similar to the fireworks explosion in the night sky published by Professor Tan Ying of Peking University in 2010, which mainly searches for the optimal solution in the solution space of the problem by generating sparks and mutations similar to the way fireworks explode and generate new fireworks.), the steps are as follows:

[0041] 1) randomly generating a population as an initial value, wherein the population includes multiple solutions;

[0042] 2) Evaluate the population through the objective function value. If the current optimal solution is better than the global optimal solution, update the global optimal solution; otherwise, increase the number of iteration records by one;

[0043] 3) Sparks are generated through fireworks explosions in the population, local optimal solutions are searched through sparks, and sparks are selected as the next generation of fireworks;

[0044] 4) Generate new fireworks through mutation. When the objective function value of the new fireworks is better than the worst objective function value in the population, replace the old fireworks corresponding to the worst objective function value with the new fireworks;

[0045] 5) If the global optimal solution is still not updated after multiple iterations, update the population through the fireworks update strategy and the population disturbance strategy, and then return to step 2) to continue iterating. When the number of iteration records reaches the threshold, execute step 6);

[0046] 6) Output the global optimal solution.

[0047] Furthermore, the fireworks update strategy refers to making local adjustments to the fireworks in the population that are closest to the optimal solution of the objective function value.

[0048] The fireworks update strategy can ensure the quality of the solution while taking into account diversity, allowing the population to iteratively evolve towards the optimal solution.

[0049] Furthermore, the population perturbation strategy refers to replacing the worst solution of the objective function value within the population with a randomly generated solution.

[0050] By perturbing the population, the population will undergo changes with a greater amplitude than explosions and mutations, increasing the possibility of the fireworks population finding the optimal solution.

[0051] The fireworks update strategy and population disturbance strategy are used to overcome the problems of high randomness and slow solution of the traditional fireworks algorithm, thereby improving the search and solution efficiency.

[0052] In addition, other algorithms known in the art may be used for planning to achieve the same technical purpose.

[0053] Furthermore, the explosion operation in step 3) includes an explosion radius, and the explosion radius is related to the objective function value corresponding to the fireworks.

[0054] Furthermore, after determining the radius of the explosion operation, it is also necessary to determine the number of sparks that each firework can produce, that is, the number of new individuals generated by the explosion. As for the number of sparks generated by the explosion, fireworks with good performance mean that there are more opportunities to generate more sparks through the explosion, which is also conducive to further making full use of the useful information obtained during the search process.

[0055] The above calculation method may result in too many or too few fireworks. In order to avoid this situation and affect the algorithm search efficiency, a piecewise function is used to limit the number of generated sparks to a certain range.

[0056] Furthermore, the number of sparks after the explosion will be greater than the number required by the firework population, and a discrimination method is needed to determine which sparks can become fireworks.

[0057] Preferably, the discrimination method is to select sparks to become fireworks in a roulette manner, so that sparks with good performance are more likely to be selected as fireworks.

[0058] Furthermore, the method for planning the split subtasks is a heuristic algorithm based on the schedulability of the split tasks. The algorithm mainly completes the task splitting and the task arrangement according to the conflicts between tasks and between tasks and resources. The detailed steps are as follows:

[0059] a. According to the split subtask set, obtain the visible time window set of each subtask;

[0060] b. After sorting the subtasks in the set of subtasks after the splitting of the splittable observation task, arrange the time windows in order;

[0061] c. After the task scheduling is completed, if there are still time windows remaining, split the tasks that were not split in step 2, and iterate the task scheduling until there are no time windows remaining or the task splitting cannot be continued. The iteration ends and the planning results are obtained.

[0062] In addition, other algorithms known in the art may also be used to achieve the same technical purpose.

[0063] Furthermore, the method for splitting the splittable tasks in step c is: when the required observation time length of an observation task exceeds, for example, twice or more, the length of the longest remaining visible time window, the task is selected to wait for splitting, and when other tasks are arranged, if there is still a remaining time window available, the task is attempted to be split. It should be pointed out that the multiple of the longest remaining visible time window length here can be adjusted according to specific needs, and can also be adjusted to, for example, 1 time, 3 times or more.

[0064] Furthermore, the calculation formula of the splittability evaluation index is:

[0065]

[0066] When a time window is known, its properties are , this time window is the longest in the visible time window for executing the task. If the task requires the observation time and the time window length to satisfy the following relationship:

[0067]

[0068] The task is divided into two subtasks: and subtasks , their properties are:

[0069]

[0070]

[0071] The parameter definitions are as described above.

[0072] After the tasks are split, the tasks can be sorted and a task execution plan can be generated based on the schedulability of the computing tasks.

[0073] Furthermore, according to the schedulability of subtasks, the method to arrange a reasonable time window for tasks is:

[0074] satellite Splittable observation tasks No. and The subtasks are scheduled simultaneously in When the following relationship is satisfied in a time window, it means there is a conflict, and the schedulability of the task needs to be calculated;

[0075]

[0076] in, Indicates satellite Splittable observation tasks No. The subtask is The start execution time within the time window, Indicates satellite Splittable observation tasks No. The subtask is The earliest actually visible time in the time window, and the other parameters are defined as described above.

[0077] Satellite for the and The calculation formula for the schedulability of a subtask is as follows:

[0078]

[0079] in, Indicates satellite Splittable observation tasks No. The subtask is The latest visible time in the time window, Indicates satellite Splittable observation tasks No. The subtask is The latest visible time in the time window, Indicates satellite Splittable observation tasks No. The subtask is The earliest visible time in the time window, Indicates satellite Splittable observation tasks No. The subtask is The earliest visible time in a time window, and the remaining parameters are defined as described above.

[0080] Furthermore, the schedulability of a satellite mission can be calculated based on the average schedulability of each subtask to be observed by the satellite.

[0081] Furthermore, the calculation method of the satellite's mission schedulability is as follows:

[0082]

[0083] After obtaining the schedulability values ​​of each subtask and satellite for the task, the tasks are sorted from high to low according to the schedulability values, and the visible time window resources are tried to be arranged for the tasks in turn. After the subtasks and time windows are selected through the arrangement of time windows, the planning process of the splittable observation task is completed, and the task execution plan of the splittable observation task is obtained, so that the task is split and planned reasonably and effectively.

[0084] On the other hand, the present invention also discloses a multi-satellite joint planning system considering task splitting, including a task receiving module, a task processing module and a task generating module.

[0085] The task receiving module is used to receive the task set to be planned and the satellite set participating in the task execution, and then pass them to the task processing module;

[0086] The task processing module is used to operate the task set and the satellite set according to steps 2 to 5 of the multi-satellite joint planning method considering task splitting, and mainly includes a splitting unit and a calculation unit:

[0087] The splitting unit divides the tasks into two categories, namely, a splittable observation task set and an indivisible observation task set, according to whether the tasks are splittable. Then, the splittable observation task set is split into a split sub-task set and a task set that does not need to be split according to a task splitting algorithm. Finally, the indivisible observation task set, the split sub-task set and the task set that does not need to be split are sent to the computing unit.

[0088] The operation unit includes a task planning model, which receives the indivisible observation task set, the split sub-task set and the unneeded task set sent by the splitting unit, obtains a planning scheme by solving the optimal solution through the task planning model, and then sends the planning scheme to the task generation module.

[0089] The task generation module is used to output the final planning solution.

[0090] On the other hand, the present invention also provides a multi-satellite joint planning device considering task splitting, which mainly includes a processor, a memory and a bus. The memory stores instructions that can be read by the processor. The processor is used to call the instructions in the memory to execute the multi-satellite joint planning method considering task splitting. The bus connects the functional components to transmit information.

[0091] By adopting the above technical solution, the present invention has the following beneficial effects:

[0092] The multi-satellite joint planning method and system considering task splitting provided by the present invention construct a task planning model and design algorithms for planning inseparable and divisible tasks respectively. The task planning model covers the planning of divisible observation tasks and inseparable observation tasks at the same time, and fully considers various constraints from satellite resource capability level, task timeliness requirements and the like. According to the characteristics of the fireworks algorithm for multi-satellite joint observation task planning problems and the shortcomings of other classical algorithms, the present invention introduces the fireworks update strategy and the population disturbance strategy. The use of these two strategies effectively improves the planning effect of the algorithm. The heuristic algorithm effectively splits the divisible observation tasks, and sorts and generates task execution plans according to the task scheduleability. After a large number of numerical experiments, it is proved that the two-stage hybrid scheduling algorithm considering task splitting proposed by the present invention provides the possibility of collaboration between multiple satellites through task splitting, and improves the task completion rate. The multi-satellite joint observation benefits brought by task splitting are also significantly improved, and the observation efficiency of the satellite is greatly improved in the end. BRIEF DESCRIPTION OF THE DRAWINGS

[0093] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0094] Figure 1 A flowchart provided for an embodiment of the present invention;

[0095] Figure 2 A schematic diagram of a detachable observation provided by an embodiment of the present invention;

[0096] Figure 3 A schematic diagram of a small-scale test set result provided by an embodiment of the present invention;

[0097] Figure 4 A schematic diagram of the results of a medium-scale test set provided by an embodiment of the present invention;

[0098] Figure 5A schematic diagram of comparison results of different planning methods provided by an embodiment of the present invention;

[0099] Figure 6 A system diagram provided for an embodiment of the present invention. DETAILED DESCRIPTION

[0100] The technical solution of the present invention will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0101] In the description of the present invention, it should be noted that the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc., indicating the orientation or positional relationship, are based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first", "second", and "third" are used for descriptive purposes only, and cannot be understood as indicating or implying relative importance.

[0102] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0103] In a preferred embodiment, Figure 1 As shown, the multi-satellite joint planning method considering task splitting provided by the present invention includes the following steps:

[0104] Step 1: Obtain the task set to be planned and the satellite set participating in the task execution;

[0105] Step 2: Divide the task set in step 1 into a divisible observation task set and an indivisible observation task set according to whether the task is divisible;

[0106] Step 3: According to the task splitting method, the splittable observation task set in step 2 is divided into a split subtask set and a task set that does not need to be split;

[0107] Step 4: Based on the input of the non-divisible observed task set, the split sub-task set and the task set that does not need to be split in steps 2 to 3, the task planning model is used to solve the optimal solution to obtain the planning solution output;

[0108] Step 5: Output the result of step 4 as the overall planning solution.

[0109] Furthermore, the task planning model in step 4 includes an objective function formula:

[0110] ;

[0111] Among them, SAT represents the satellite set, i represents the i-th satellite;

[0112] Represents a set of splittable observation tasks, subtasks;

[0113] Indicates time window resources, total Visible time windows;

[0114] is the decision variable, indicating the satellite Indivisible observation tasks or tasks that do not need to be split Arranged in time window, when the task is successfully scheduled ,otherwise ;

[0115] is the decision variable, indicating the satellite Splittable observation tasks Subtasks Arranged in time window, when the subtask is successfully scheduled ,otherwise ;

[0116] It represents the benefits of observation tasks that cannot be split or tasks that do not need to be split;

[0117] The constraints of the objective function formula are as follows:

[0118] (1) The observation task cannot be split or does not need to be split, and the actual start time is later than the earliest allowed execution time: ;

[0119] in, Indicates the earliest allowed start time of the satellite observation mission; Indicates satellite The observation mission The start time of execution; Indicates an observation task that cannot be split (if there is a task that does not need to be split, it is also included), represents a set of inseparable observation tasks (if there are tasks that do not need to be split, they are also included), with a total of tasks;

[0120] (2) For observation tasks that cannot be split or do not need to be split, the actual end time must be earlier than the latest allowed end time: ;

[0121] in, Indicates the actual observation time of the observation task; Indicates the latest allowed end time of the satellite observation mission;

[0122] (3) The actual start time of the subtask of the splittable observation task is later than the earliest allowed execution time of the splittable observation task:

[0123] ;

[0124] in, Indicates satellite Splittable observation tasks The earliest allowed start time, Indicates satellite Splittable observation tasks Subtasks The start time of execution, Represents a splittable observation task, which is split into subtasks;

[0125] (4) The actual end time of the subtask of the splittable observation task must be earlier than the latest allowed execution time of the splittable observation task:

[0126] ;

[0127] in, Indicates satellite Splittable observation tasks Subtasks The actual observation time; Indicates satellite Splittable observation tasks The latest running end time;

[0128] (5) Observation tasks that cannot be split or do not need to be split need to be started within the visible time window of the observation target:

[0129] ;

[0130] in, represents the earliest visible time of the kth time window;

[0131] (6) Observation tasks cannot be split or do not need to be split, and need to be completed within the visible time window of the observation target:

[0132] ;

[0133] in, Indicates the latest visible time of the kth time window;

[0134] (7) Any subtask of the splittable observation task needs to be started within the visible time window of the observation target:

[0135] ;

[0136] (8) Any subtask of the splittable observation task needs to be completed within the visible time window of the observation target:

[0137] ;

[0138] (9) The observation task cannot be split or does not need to be split. The actual observation time must be equal to the required observation time:

[0139] ;in, Indicates the observation time required by the observation task;

[0140] (10) The sum of the actual observation time of each subtask of the divisible observation task must be equal to the required observation time of the divisible observation task:

[0141] ;

[0142] (11) Any two observation tasks that cannot be split or do not need to be split must meet the minimum satellite observation interval requirement:

[0143] ;

[0144] in, Indicates satellite The observation mission The start time of execution; Indicates the minimum satellite observation interval;

[0145] (12) The execution of any two subtasks of the separable observation task meets the minimum satellite observation interval requirement:

[0146] ;

[0147] in, Indicates satellite Splittable observation tasks No. The start time of each subtask;

[0148] (13) Any observation task that cannot be split or does not need to be split, and the subtasks of the splittable observation task meet the minimum satellite observation interval requirements:

[0149] ;

[0150] ;

[0151] ;

[0152] (14) Each indivisible observation task or task that does not need to be split is executed at most once:

[0153] ;

[0154] (15) Each subtask of a separable observation task is executed at most once:

[0155] .

[0156] Among them, constraints (1)-(4) describe that task execution must meet the time requirements of the task itself; constraints (5)-(8) describe that task execution must ensure visibility to the target and start and end within the visible time window; constraints (9)-(10) describe that the task execution time must be consistent with the required duration; constraints (11)-(13) describe the restrictions on task scheduling due to the minimum time interval; and constraints (14)-(15) describe the restrictions on the number of task executions.

[0157] Furthermore, the method for solving the task planning model in step 4 is:

[0158] In the first stage, the inseparable observation tasks and the tasks that do not need to be split are directly planned to obtain planning schemes for the inseparable observation tasks and the tasks that do not need to be split, and the remaining time windows after the inseparable observation tasks and the tasks that do not need to be split are calculated;

[0159] In the second stage, the split subtasks are planned to obtain the planning scheme of the split observation task;

[0160] Finally, the planning schemes for the inseparable observation tasks and the tasks that do not need to be split are sent to step 5 together with the planning scheme for the splittable observation tasks as the overall planning scheme output.

[0161] Furthermore, the method for planning the inseparable tasks and the tasks that do not need to be split (improved fireworks algorithm) is as follows:

[0162] 1) randomly generating a population as an initial value, wherein the population includes multiple solutions;

[0163] 2) Evaluate the population through the objective function value. If the current optimal solution is better than the global optimal solution, update the global optimal solution; otherwise, increase the number of iteration records by one;

[0164] 3) Sparks are generated through fireworks explosions in the population, local optimal solutions are searched through sparks, and sparks are selected as the next generation of fireworks;

[0165] 4) Generate new fireworks through mutation. When the objective function value of the new fireworks is better than the worst objective function value in the population, replace the old fireworks corresponding to the worst objective function value with the new fireworks;

[0166] 5) If the global optimal solution is still not updated after multiple iterations, update the population through the fireworks update strategy and the population disturbance strategy, and then return to step 2) to continue iterating. When the number of iteration records reaches the threshold, execute step 6);

[0167] 6) Output the global optimal solution.

[0168] In addition, in other embodiments, other algorithms known in the art may be used for planning to achieve the same technical purpose.

[0169] Furthermore, the fireworks update strategy refers to making local adjustments to the fireworks in the population that are closest to the optimal solution of the objective function value.

[0170] The fireworks update strategy can ensure the quality of the solution while taking into account diversity, allowing the population to iteratively evolve towards the optimal solution.

[0171] Furthermore, the population perturbation strategy refers to replacing the worst solution of the objective function value within the population with a randomly generated solution.

[0172] By perturbing the population, the population will undergo changes with a greater amplitude than explosions and mutations, increasing the possibility of the fireworks population finding the optimal solution.

[0173] Furthermore, in the improved fireworks algorithm, the explosion operation in step 3) includes an explosion radius, and the explosion radius is related to the objective function value corresponding to the fireworks. The calculation formula of the explosion radius is as follows:

[0174]

[0175] in, Represents the preset explosion radius adjustment constant; represents the minimum value of the objective function in the current population, where the objective function refers to the objective function in the model; Represents a very small quantity to avoid division by zero errors in formulas; represents the fitness value; i represents the individual number; P represents the population.

[0176] After determining the radius of the explosion operation, it is also necessary to determine the number of sparks that each firework can produce, that is, the number of new individuals generated by the explosion. For the number of sparks generated by the explosion, a firework with good performance means more opportunities to generate more sparks through the explosion, which is also conducive to further making full use of the useful information obtained during the search process. The method for calculating the number of sparks generated by the explosion is as follows:

[0177]

[0178] in, Indicates the spark quantity adjustment constant, represents the maximum value of the objective function in the current population, Represents a very small quantity and avoids division by zero errors in formulas.

[0179] The above calculation method may result in too many or too few fireworks. In order to avoid this situation and affect the algorithm search efficiency, this embodiment uses a piecewise function to limit the number of generated sparks to a certain range. The calculation formula is as follows:

[0180]

[0181] in, represents the rounding function, are two preset constants.

[0182] The number of sparks after the explosion will be more than the number required by the firework population, and a discrimination method is needed to determine which sparks can become fireworks.

[0183] Furthermore, the discrimination method refers to selecting sparks to become fireworks in a roulette manner, so that sparks with good performance are more likely to be selected as fireworks.

[0184] Furthermore, if Figure 2 As shown in the figure, the method for planning the split subtasks is a heuristic algorithm based on the schedulability of the split tasks. The algorithm mainly completes the task splitting and the task arrangement according to the conflicts between tasks and between tasks and resources. The detailed steps are as follows:

[0185] a. According to the split subtask set, obtain the visible time window set of each subtask;

[0186] b. After sorting the subtasks in the split subtask set according to their schedulability, arrange the time windows in sequence;

[0187] c. After the task scheduling is completed, if there are still time windows remaining, split the tasks that were not split in step 2, and iterate the task scheduling until there are no time windows remaining or the task splitting cannot be continued. The iteration ends and the planning results are obtained.

[0188] In addition, in other embodiments, other algorithms known in the art may also be used to achieve the same technical purpose.

[0189] Furthermore, the method for splitting the separable tasks in step c is: when the required observation time length of an observation task exceeds twice or more the length of the longest remaining visible time window, the task is selected and waits for splitting. When other tasks are arranged, if there is still a remaining time window available, the task is attempted to be split.

[0190] Furthermore, the calculation formula of the splittability evaluation index is:

[0191]

[0192] When a time window is known, its properties are , this time window is the longest in the visible time window for executing the task. If the task requires the observation time and the time window length to satisfy the following relationship:

[0193]

[0194] The task is divided into two subtasks: and subtasks , their properties are:

[0195]

[0196]

[0197] The parameter definitions are as described above.

[0198] After the tasks are split, the tasks can be sorted and a task execution plan can be generated based on the schedulability of the computing tasks.

[0199] Furthermore, according to the schedulability of subtasks, the method to arrange a reasonable time window for tasks is:

[0200] satellite Splittable observation tasks No. and The subtasks are scheduled simultaneously in When the following relationship is satisfied in a time window, it means there is a conflict, and the schedulability of the task needs to be calculated;

[0201]

[0202] in, Indicates satellite Splittable observation tasks No. The subtask is The start execution time within the time window, Indicates satellite Splittable observation tasks No. The subtask is The earliest actually visible time in the time window, and the other parameters are defined as described above.

[0203] Satellite for the and The calculation formula for the schedulability of a subtask is as follows:

[0204]

[0205] in, Indicates satellite Splittable observation tasks No. The subtask is The latest visible time in the time window, Indicates satellite Splittable observation tasks No. The subtask is The latest visible time in the time window, Indicates satellite Splittable observation tasks No. The subtask is The earliest visible time in the time window, Indicates satellite Splittable observation tasks No. The subtask is The earliest visible time in a time window, and the remaining parameters are defined as described above.

[0206] Furthermore, the schedulability of a satellite mission can be calculated based on the average schedulability of each subtask to be observed by the satellite.

[0207] Furthermore, the calculation method of the satellite's mission schedulability is as follows:

[0208]

[0209] After obtaining the schedulability values ​​of each subtask and satellite for the task, the tasks are sorted from high to low according to the schedulability values, and visible time window resources are tried to be arranged for the tasks in turn. After completing the selection of subtasks and time windows through task scheduling, the planning process of the divisible observation task is completed, and the task execution plan of the divisible observation task is obtained.

[0210] In a more specific embodiment, we compare the experimental results in the following environment: the operating environment is Windows 10, the coding software is Matlab 2020a, the experimental test set is generated by STK 11.2 software, and the test set includes a test set containing only inseparable observation tasks and a test set containing both inseparable observation tasks and splittable observation tasks. The test set containing only inseparable observation tasks is used to verify the optimization effect of the evolutionary algorithm proposed in the first stage. The test set containing both inseparable observation tasks and splittable observation tasks can simultaneously test the planning effect of the evolutionary algorithm on the inseparable observation task sequence and the planning effect of the heuristic algorithm on the splittable observation task sequence.

[0211] In order to increase the fairness of algorithm experiments, the present invention uses three algorithms for solving satellite mission planning as benchmark algorithms for comparison. The present invention selects three algorithms: genetic algorithm (IGA), adaptive large neighborhood search algorithm (ALNS-I), and neighborhood search algorithm (LS). The adaptive large neighborhood search algorithm (ALNS-I) uses neighborhood construction and destruction to continuously obtain new neighborhood structures to achieve iterative search. The neighborhood search algorithm (LS) finds the optimal neighborhood structure by iterative insertion.

[0212] Since the evaluation mechanism of the algorithm is also important for evaluating the optimization effect of the algorithm, in order to effectively evaluate the optimization performance of the evolutionary algorithm, this comparative experiment sets the maximum value ( ),average value( ), minimum value( ) and the standard deviation of the optimization results ( ) as the evaluation index.

[0213] The comparative experiment contents are as follows:

[0214] First, the performance of the improved fireworks algorithm in solving the inseparable observation task planning is verified. The inseparable observation task set contains three test sets of large, medium and small sizes, which are represented by _Y, corresponding to the number of tasks in the test set, and Y represents the internal sequence number under each test set size:

[0215] 1. Experiments on a small-scale test set. The results are shown in Table 1.

[0216] Table 1 Small-scale test set results:

[0217]

[0218] As can be seen from Table 1, the improved fireworks algorithm (FWA-I) can find the optimal value in each search process for the test set of 100 tasks, and the genetic algorithm (IGA) and the adaptive large neighborhood search algorithm (ALNS-I) can also find the optimal value in each search in the 100_2 test set. The average fluctuation of the improved fireworks algorithm (FWA-I) is better than that of the genetic algorithm (IGA) and the adaptive large neighborhood search algorithm (ALNS-I), and the fluctuation of the neighborhood search algorithm (LS) is the most obvious. In the process of increasing the number of tasks from 200 to 300, the observed benefits of the task sequence did not increase rapidly accordingly, and the growth rate tended to be flat.

[0219] 2. Experiments on a medium-sized test set. The results are shown in Table 2.

[0220] Table 2 Results of medium-sized test set:

[0221]

[0222] The number of satellites in the medium-sized test set increases as the number of missions increases, which makes it possible for more missions to be successfully scheduled. It can be clearly seen from the results that the optimization stability of the improved fireworks algorithm does not deteriorate with the increase in mission size, but the other comparison algorithms have produced obvious fluctuations. Overall, the improved fireworks algorithm performs better in terms of stability than the genetic algorithm and the adaptive large neighborhood search algorithm.

[0223] 3. Large-scale test set test, the results are shown in Table 3. In the large-scale test set, the planning benefit obtained by the improved fireworks algorithm proposed in the present invention has obvious advantages over the benchmark algorithm. The benefit values ​​obtained by the test sets containing 900 tasks and 1000 tasks respectively are within a certain range, which shows that the available time window resources are limited, and the tasks that can be arranged while satisfying various constraints are limited. The performance of the genetic algorithm and the adaptive large neighborhood search algorithm is poorer than that of the improved fireworks algorithm, but is significantly better than the neighborhood search algorithm. The neighborhood search algorithm may fall into a local optimal situation during the optimization process of a large-scale test set, resulting in poor optimization effect.

[0224] Table 3 Large-scale test set results:

[0225]

[0226] Then, the performance of the heuristic algorithm in solving the splittable observation task planning is verified. The visible time window resources used by the heuristic algorithm are the remaining available resources after the previous part of the non-splittable task planning. 60 and 120 splittable observation tasks are set in the small-scale test set and the medium-scale test set respectively. The experimental results are as follows: Figure 3 and Figure 4 shown.

[0227] Figure 3 100% represents the planning benefit that can be obtained after using the heuristic algorithm. The experimental results show that after using the heuristic algorithm, the task is split into subtasks and some tasks are successfully scheduled after sorting. The splitting process enables the algorithm to find available time windows from the remaining time window resources to complete task splitting and successful planning.

[0228] like Figure 3 As shown in the figure, the smaller the number of indivisible observation tasks, the more obvious the benefits of planning divisible observation tasks are, which is related to the fact that indivisible observation tasks occupy less time window resources. When the scale of indivisible tasks reaches a certain level, it becomes difficult to successfully plan tasks by splitting them.

[0229] like Figure 4 As shown in the figure, by comparing the results of 120 divisible observation tasks, it can be found that the increase in task scale has reduced the benefit improvement effect of planning divisible tasks, but it still maintains a relatively significant benefit performance.

[0230] Finally, the present invention also solves the planning results of using task splitting and not using the task splitting algorithm, respectively, using the two-stage hybrid scheduling algorithm (FWA-I-HA) and using the improved fireworks algorithm (FWA-I) alone to solve the test set. Using the improved fireworks algorithm (FWA-I) alone means using the algorithm to process both splittable observation tasks and non-splittable observation tasks. The number of non-splittable observation tasks in the test set are 100, 200, 300, 400, 500 and 600, respectively. When the number of non-splittable tasks is in the range of 100-300, the number of splittable observation tasks is 60, and when the number of non-splittable tasks is in the range of 400-600, the number of splittable observation tasks is 120. The experimental results are as follows. Figure 5 As shown, it effectively proves that the two-stage hybrid scheduling method is better than using only the fireworks algorithm as the planning method, and also verifies the advantages of the observation task planning method based on task splitting.

[0231] like Figure 5As shown in the figure, the two-stage hybrid scheduling algorithm considering task splitting performs better in planning when there are more indivisible and divisible observation tasks. The two-stage approach can more easily bring out the strengths of the improved fireworks algorithm and the heuristic algorithm in task planning. However, it is difficult to obtain satisfactory results if only the improved fireworks algorithm is used to process two types of tasks at the same time.

[0232] Based on the above experiments, the two-stage hybrid scheduling algorithm (FWA-I-HA) proposed in this paper can effectively solve the multi-satellite joint observation task planning problem. The improved fireworks algorithm (FWA-I) performs significantly better than the benchmark algorithm used to compare the algorithm effect in the planning of inseparable observation tasks. The heuristic algorithm (HA) splits the tasks and arranges them in a heuristic way, which improves the task completion rate and further improves the operating efficiency of the satellite system.

[0233] The joint earth observation of multiple satellites not only improves the completion rate of the task but also gives full play to the role of the satellite system. In order to effectively cope with the challenges brought by multi-satellite observation, the present invention designs a two-stage hybrid scheduling algorithm for planning inseparable and divisible tasks respectively. The two-stage hybrid scheduling algorithm covers the planning of divisible observation tasks and inseparable observation tasks at the same time, and fully considers various constraints such as resource capacity level and task timeliness requirements. The present invention introduces the fireworks update strategy and the population disturbance strategy to the fireworks algorithm according to the characteristics of the multi-satellite joint observation task planning problem and the shortcomings of the classical algorithm. The use of the two strategies effectively improves the planning effect of the algorithm. The heuristic algorithm realizes effective task splitting for the divisible observation task, and generates a task execution plan according to the task schedulability sorting. After a large number of numerical experiments, it can be seen that the two-stage hybrid scheduling algorithm considering task splitting proposed by the present invention increases the possibility of collaboration between multiple satellites by task splitting, and improves the task completion rate. The improvement of the benefits of multi-satellite joint observation brought about by task splitting is obvious, which can fully improve the observation efficiency of satellites.

[0234] In another specific implementation, this embodiment also provides a multi-satellite joint planning system considering task splitting, including a task receiving module, a task processing module and a task generating module, such as Figure 6 shown.

[0235] The task receiving module is used to receive the task set to be planned and the satellite set participating in the task execution, and then pass them to the task processing module;

[0236] The task processing module is used to operate the task set and the satellite set according to steps 2 to 5 of the multi-satellite joint planning method considering task splitting, and mainly includes a splitting unit and a computing unit:

[0237] The splitting unit divides the tasks into two categories, namely, a splittable observation task set and an indivisible observation task set, according to whether the tasks are splittable. Then, the splittable observation task set is split into a split sub-task set and a task set that does not need to be split according to a task splitting algorithm. Finally, the indivisible observation task set, the split sub-task set and the task set that does not need to be split are sent to the computing unit.

[0238] The operation unit includes a task planning model, which receives the indivisible observation task set, the split sub-task set and the unneeded task set sent by the splitting unit, obtains the optimal planning scheme by solving the optimal solution through the task planning model, and then sends the optimal planning scheme to the task generation module.

[0239] The mission planning model includes the objective function formula:

[0240]

[0241] Among them, SAT represents the satellite set, i represents the i-th satellite;

[0242] Represents a set of splittable observation tasks, subtasks;

[0243] Indicates time window resources, total Visible time windows;

[0244] is the decision variable, indicating the satellite Indivisible observation tasks or tasks that do not need to be split Arranged in time window, when the task is successfully scheduled ,otherwise ;

[0245] is the decision variable, indicating the satellite Splittable observation tasks Subtasks Arranged in time window, when the subtask is successfully scheduled ,otherwise ;

[0246] It represents the benefits of observation tasks that cannot be split or tasks that do not need to be split; represents the benefit obtained from the splittable observation task;

[0247] The constraints of the objective function formula are as follows:

[0248] (1) The observation task cannot be split or does not need to be split, and the actual start time is later than the earliest allowed execution time: ;

[0249] in, Indicates the earliest allowed start time of the satellite observation mission; Indicates satellite The observation mission The start time of execution; Indicates an observation task that cannot be split (if there is a task that does not need to be split, it is also included), Represents a set of observation tasks that cannot be split (if there are tasks that do not need to be split, they are also included), totaling tasks;

[0250] (2) For observation tasks that cannot be split or do not need to be split, the actual end time must be earlier than the latest allowed end time: ;

[0251] in, Indicates the actual observation time of the observation task; Indicates the latest allowed end time of the satellite observation mission;

[0252] (3) The actual start time of the subtask of the splittable observation task is later than the earliest allowed execution time of the splittable observation task:

[0253] ;

[0254] in, Indicates satellite Splittable observation tasks The earliest allowed start time, Indicates satellite Splittable observation tasks Subtasks The start time of execution, Represents a splittable observation task, which is split into subtasks;

[0255] (4) The actual end time of the subtask of the splittable observation task must be earlier than the latest allowed execution time of the splittable observation task:

[0256] ;

[0257] in, Indicates satellite Splittable observation tasks Subtasks The actual observation time; Indicates satellite Splittable observation tasks The latest running end time;

[0258] (5) Observation tasks that cannot be split or do not need to be split need to be started within the visible time window of the observation target:

[0259] ;

[0260] in, represents the earliest visible time of the kth time window;

[0261] (6) Observation tasks cannot be split or do not need to be split, and need to be completed within the visible time window of the observation target:

[0262] ;

[0263] in, Indicates the latest visible time of the kth time window;

[0264] (7) Any subtask of the splittable observation task needs to be started within the visible time window of the observation target:

[0265] ;

[0266] (8) Any subtask of the splittable observation task needs to be completed within the visible time window of the observation target:

[0267] ;

[0268] (9) The observation task cannot be split or does not need to be split. The actual observation time must be equal to the required observation time:

[0269] ;in, Indicates the observation time required by the observation task;

[0270] (10) The sum of the actual observation time of each subtask of the divisible observation task must be equal to the required observation time of the divisible observation task:

[0271] ;

[0272] (11) Any two observation tasks that cannot be split or do not need to be split must meet the minimum satellite observation interval requirement:

[0273] ;

[0274] in, Indicates satellite The observation mission The start time of execution; Indicates the minimum satellite observation interval;

[0275] (12) The execution of any two subtasks of the separable observation task meets the minimum satellite observation interval requirement:

[0276] ;

[0277] in, Indicates satellite Splittable observation tasks No. The start time of each subtask;

[0278] (13) Any observation task that cannot be split or does not need to be split, and the subtasks of the splittable observation task meet the minimum satellite observation interval requirements:

[0279] ;

[0280] ;

[0281] ;

[0282] (14) Each indivisible observation task or task that does not need to be split is executed at most once:

[0283] ;

[0284] (15) Each subtask of a separable observation task is executed at most once:

[0285] .

[0286] Among them, constraints (1)-(4) describe that task execution must meet the time requirements of the task itself; constraints (5)-(8) describe that task execution must ensure visibility to the target and start and end within the visible time window; constraints (9)-(10) describe that the task execution time must be consistent with the required duration; constraints (11)-(13) describe the restrictions on task scheduling due to the minimum time interval; and constraints (14)-(15) describe the restrictions on the number of task executions.

[0287] The method for solving the task planning model is:

[0288] In the first stage, the inseparable observation tasks and the tasks that do not need to be split are directly planned to obtain planning schemes for the inseparable observation tasks and the tasks that do not need to be split, and the remaining time windows after the inseparable observation tasks and the tasks that do not need to be split are calculated;

[0289] In the second stage, the split subtasks are planned to obtain the planning scheme of the split observation task;

[0290] Finally, the planning schemes for the inseparable observation tasks and the tasks that do not need to be split are sent to the task generation module together with the planning scheme for the splittable observation tasks as the overall planning scheme output.

[0291] The method for planning the tasks that cannot be split and do not need to be split (improved fireworks algorithm) is as follows:

[0292] 1) randomly generating a population as an initial value, wherein the population includes multiple solutions;

[0293] 2) Evaluate the population through the objective function value. If the current optimal solution is better than the global optimal solution, update the global optimal solution; otherwise, increase the number of iteration records by one;

[0294] 3) Sparks are generated through fireworks explosions in the population, local optimal solutions are searched through sparks, and sparks are selected as the next generation of fireworks;

[0295] 4) Generate new fireworks through mutation. When the objective function value of the new fireworks is better than the worst objective function value in the population, replace the old fireworks corresponding to the worst objective function value with the new fireworks;

[0296] 5) If the global optimal solution is still not updated after multiple iterations, update the population through the fireworks update strategy and the population disturbance strategy, and then return to step 2) to continue iterating. When the number of iteration records reaches the threshold, execute step 6);

[0297] 6) Output the global optimal solution.

[0298] The method for planning the subtasks after splitting is a heuristic algorithm based on the schedulability of the tasks after splitting. The algorithm mainly completes the task splitting and the task arrangement according to the conflicts between tasks and between tasks and resources. The detailed steps are as follows:

[0299] a. According to the split subtask set, obtain the visible time window set of each subtask;

[0300] b. After sorting the subtasks in the set of subtasks after the splitting of the splittable observation task, arrange the time windows in order;

[0301] c. After the task scheduling is completed, if there are still time windows remaining, split the unsplit tasks in the first round, and iterate the task scheduling until there are no time windows remaining or the task splitting cannot be continued. The iteration ends and the planning results are obtained.

[0302] The task generation module is used to output the final planning solution.

[0303] In addition, in another embodiment, the solution of the present invention can also be implemented by a multi-satellite joint planning device considering task splitting, which mainly includes a processor, a memory and a bus;

[0304] The memory stores instructions that can be read by the processor, and the size is flexibly configured according to storage needs;

[0305] The bus connects the functional components of the computer to transmit information;

[0306] The processor is used to call the instructions in the memory to execute a multi-satellite joint planning method considering task splitting.

[0307] In a preferred embodiment, preferably, the system configuration information may be as follows:

[0308] CPU: Core i7-7700 3.6 GHz; Memory: 8GB.

[0309] In another embodiment, the present solution can be implemented by a device, which may include a corresponding module for performing each or several steps in the above-mentioned embodiments. The module may be one or more hardware modules specially configured to perform the corresponding steps, or implemented by a processor configured to perform the corresponding steps, or stored in a computer-readable medium for implementation by a processor, or implemented by some combination.

[0310] The processor performs the various methods and processes described above. For example, the method implementation in the present solution can be implemented as a software program, which is tangibly contained in a machine-readable medium, such as a memory. In some embodiments, part or all of the software program can be loaded and / or installed via a memory and / or a communication interface. When the software program is loaded into the memory and executed by the processor, one or more steps in the method described above can be executed. Alternatively, in other embodiments, the processor can be configured to perform one of the above methods by any other appropriate means (for example, by means of firmware).

[0311] The device can be implemented using a bus architecture. The bus architecture can include any number of interconnecting buses and bridges, depending on the specific application and overall design constraints of the hardware. The bus connects various circuits including one or more processors, memories, and / or hardware modules together. The bus can also connect various other circuits such as peripherals, voltage regulators, power management circuits, external antennas, etc.

[0312] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Component (EISA) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc.

[0313] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. 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 replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A multi-satellite joint planning method considering task splitting, characterized in that: The steps include: Step 1: Obtain the task set to be planned and the satellite set participating in the task execution; Step 2: Divide the task set in step 1 into a divisible observation task set and an indivisible observation task set according to whether the task is divisible; Step 3: According to the task splitting method, the splittable observation task set in step 2 is divided into a split subtask set and a task set that does not need to be split; Step 4: Based on the input of the non-divisible observed task set, the split sub-task set and the task set that does not need to be split in steps 2 to 3, the task planning model is used to solve the optimal solution to obtain the planning solution output; The mission planning model includes the objective function formula: ； in, represents the satellite set, Indicates Satellites; Represents a set of observation tasks that cannot be split and tasks that do not need to be split; Represents a set of splittable observation tasks, subtasks; Indicates time window resources, total Visible time windows; is the decision variable, indicating the satellite Indivisible observation tasks or tasks that do not need to be split Arranged in time window, when the task is successfully scheduled ,otherwise ; is the decision variable, indicating the satellite Splittable observation tasks Subtasks Arranged in time window, when the subtask is successfully scheduled ,otherwise ; It represents the benefits of observation tasks that cannot be split or do not need to be split; represents the benefit obtained from the splittable observation task; The method for solving the task planning model in step 4 is: In the first stage, the inseparable observation tasks and the tasks that do not need to be split are directly planned to obtain planning schemes for the inseparable observation tasks and the tasks that do not need to be split, and the remaining time windows after the inseparable observation tasks and the tasks that do not need to be split are calculated; In the second stage, the split subtasks are planned to obtain the planning scheme of the split observation task; Finally, the planning schemes for the inseparable observation tasks and the tasks that do not need to be split are sent to step 5 together with the planning schemes for the splittable observation tasks as the overall planning scheme output; In the first stage, the method for planning the inseparable observation tasks and the tasks that do not need to be split is as follows: 1) randomly generating a population as an initial value, wherein the population includes multiple solutions; 2) Evaluate the population through the objective function value of the task planning model. If the current optimal solution is better than the global optimal solution, update the global optimal solution, otherwise increase the number of iteration records by one; 3) Sparks are generated through fireworks explosions in the population, local optimal solutions are searched through sparks, and sparks are selected as the next generation of fireworks; 4) Generate new fireworks through mutation. When the objective function value of the new fireworks is better than the worst objective function value in the population, replace the old fireworks corresponding to the worst objective function value with the new fireworks; 5) If the global optimal solution is still not updated after multiple iterations, update the population through the fireworks update strategy and the population disturbance strategy, and then return to step 2) to continue iterating. When the number of iteration records reaches the threshold, execute step 6); 6) Output the global optimal solution; In the second stage, the method for planning the split subtasks is as follows: a. According to the split subtask set, obtain the visible time window set of each subtask; b. After sorting the subtasks in the set of subtasks after the splitting of the splittable observation task, arrange the time windows in order; The schedulability of subtasks specifically includes: Satellite Splittable observation tasks No. and The subtasks are scheduled simultaneously in When the following relationship is satisfied in a time window, it means there is a conflict, and the schedulability of the task needs to be calculated; ； in, Indicates satellite Splittable observation tasks No. The subtask is The start execution time within the time window, Indicates satellite Splittable observation tasks No. The subtask is The earliest actually visible time within a time window; Indicates the minimum satellite observation interval; Indicates satellite Splittable observation tasks Subtasks The actual observation time; Satellite for the and The schedulability of subtasks The calculation formula is as follows: ； in, Indicates satellite Splittable observation tasks No. The subtask is The latest visible time in the time window, Indicates satellite Splittable observation tasks No. The subtask is The latest visible time in the time window, Indicates satellite Splittable observation tasks No. The subtask is The earliest visible time in the time window, Indicates satellite Splittable observation tasks No. The subtask is The earliest visible time in a time window; c. After the task scheduling is completed, if there are still time windows remaining, split the tasks that were not split in step 2, and iterate the task scheduling until there are no time windows remaining or the task splitting cannot be continued. The iteration ends and the planning results are obtained; specifically, when the required observation time length of an observation task exceeds twice or more the length of the longest remaining visible time window, the task is selected and waits for splitting. When other tasks are scheduled, if there are still remaining time windows available, try to split the task; Separability evaluation index The calculation formula is: ； in, Indicates the observation time required by the observation task; Indicates The latest visible time of the time window; Indicates The earliest visible time of a time window; Step 5: Output the result of step 4 as the overall planning solution.

2. The multi-satellite joint planning method considering task splitting as claimed in claim 1, characterized in that: The constraints of the objective function include: (1) The observation task cannot be split or does not need to be split, and the actual start time is later than the earliest allowed execution time; (2) For observation tasks that cannot be split or do not need to be split, the actual end time must be earlier than the latest allowed end time; (3) The actual start time of the subtask of the splittable observation task is later than the earliest allowed execution time of the splittable observation task; (4) The actual end time of the subtask of the splittable observation task must be earlier than the latest allowed execution time of the splittable observation task; (5) The observation task cannot be split or does not need to be split, and needs to be started within the visible time window of the observation target; (6) The observation task cannot be split or does not need to be split, and the task needs to be completed within the visible time window of the observation target; (7) Any subtask of a splittable observation task needs to be started within the visible time window of the observation target; (8) Any subtask of the splittable observation task needs to be completed within the visible time window of the observation target; (9) The observation task cannot be split or does not need to be split, and the actual observation time must be equal to the required observation time; (10) The sum of the actual observation time of each subtask of the divisible observation task must be equal to the required observation time of the divisible observation task; (11) Any two observation tasks that cannot be split or do not need to be split must meet the minimum satellite observation interval requirements; (12) The execution of any two subtasks of the separable observation task meets the minimum observation interval requirement of the satellite; (13) Any observation task that cannot be split or does not need to be split, and the subtask of the observation task that can be split, need to ensure the minimum satellite observation interval requirement; (14) Each indivisible observation task or task that does not need to be split is executed at most once; (15) Each subtask of a separable observation task is executed at most once.

3. The multi-satellite joint planning method considering task splitting as claimed in claim 1, characterized in that: The fireworks update strategy refers to local adjustment of the fireworks in the population that are closest to the optimal solution of the objective function value; the population disturbance strategy refers to replacing the worst solution of the objective function value in the population with a randomly generated solution.

4. A multi-satellite joint planning system considering task splitting using the method as described in any one of claims 1 to 3, characterized in that: Including task receiving module, task processing module and task generating module: The task receiving module is used to receive the task set to be planned and the satellite set participating in the task execution, and then pass them to the task processing module; The task processing module is used to operate the task set and the satellite set, and includes a splitting unit and a computing unit: The splitting unit first divides the task into a splittable observation task set and an indivisible observation task set according to whether the task is splittable, then splits the splittable observation task set into a split subtask set and a task set that does not need to be split, and finally sends the indivisible observation task set, the split subtask set and the task set that does not need to be split to the computing unit; The computing unit includes a task planning model, receives the inseparable observation task set, the split subtask set and the unneeded split task set sent by the splitting unit, obtains a planning scheme by solving the optimal solution through the task planning model, and then sends the planning scheme to the task generation module; The task generation module is used to output the final planning solution.

Images