A target aggregation method considering the imaging mode differences of earth observation payloads

Abstract

The present invention relates to the technical field of remote sensing satellites, and specifically to a target aggregation method considering the imaging mode differences of earth observation payloads, comprising the following steps: S1: Determine whether the aggregation observation conditions are met for multiple point target observations; S2: In the case where the aggregation observation conditions are met, provide an initialized task input to the resource scheduling system, and generate an aggregation plan through a pre-aggregation algorithm; S3: Dynamically adjust the aggregation plan through an in-process aggregation algorithm according to the actual task satisfaction situation during the resource scheduling process. Based on the conventional pre-static aggregation, the present invention attempts to integrate the aggregation task into the resource scheduling process. According to the distribution of unfulfilled tasks in the original scheduling plan and the currently generated planned plan, a new aggregation task is generated, and after evaluation, it is handed over to the scheduling system for processing, that is, in-process dynamic aggregation is achieved, thereby enhancing the adequacy and effectiveness of the task aggregation operation.

Description

Technical Field

[0001] The present invention relates to the technical field of remote sensing satellites, and specifically to a target aggregation method considering the imaging mode differences of earth observation payloads. Background Art

[0002] When current remote sensing satellites are applied, the imaging area is usually regarded as a single target, and the method of observing one by one is adopted, which has deficiencies such as low utilization efficiency of satellite resources and low mission satisfaction rate. According to the imaging mechanism of satellite payloads, if the observation angles of multiple point targets are within a certain range and there is an intersection with sufficient length (or the condition for splicing) in their imaging times, aggregated observation can be considered, as shown in the appendix. Figure 1 Scholars at home and abroad are researching efficient observation methods for dense point targets. Cohen analyzed the possibility of simultaneous coverage of multiple point targets by a single strip, Wang Jun et al. considered the problem of the overlap degree of time windows during the aggregated observation of adjacent point targets, and Xu Yula et al. analyzed the problem of task merging and observation under static conditions. However, these studies are only limited to being realized under the premise of a fixed observation angle, and do not consider the impact of satellite payloads in different imaging modes on the possibility of aggregated observation of targets. To address the above deficiencies, this technology improves the aggregated constraint condition model and designs an optimization algorithm for two aggregation stages. Summary of the Invention

[0003] The present invention provides a target aggregation method considering the imaging mode differences of earth observation payloads to solve the problems raised in the above background art.

[0004] To achieve the above object, the present invention provides the following technical solutions:

[0005] A target aggregation method considering the imaging mode differences of earth observation payloads, comprising the following steps:

[0006] S1: Determine whether the aggregated observation condition is met for multiple point target observations;

[0007] S2: When the aggregated observation condition is met, provide an initial task input to the resource scheduling system, and generate an aggregation plan through a pre-aggregation algorithm;

[0008] S3: Dynamically adjust the aggregation plan through an in-process aggregation algorithm according to the actual task satisfaction situation during the resource scheduling process.

[0009] Preferably, step S1 specifically includes:

[0010] Let the earth observation satellite imaging requirement for point target the t th visible window be , and They are the start time and end time of the window, and is the payload roll angle corresponding to the visible window;

[0011] In the observation mode of fixed-strip imaging, for any two fixed-strip imaging tasks and the aggregation observation conditions are:

[0012] (1);

[0013] Among them, is for the set of visible windows of; is the minimum working time required for single imaging; is the instantaneous field of view angle of the payload.

[0014] Preferably, when the fixed-strip imaging task on and on meet the aggregation observation conditions, the aggregation task is denoted as , and its characteristic parameters include the visible window and the roll angle interval ;

[0015] Among them, the calculation method of is:

[0016] (2);

[0017] the calculation method of is:

[0018] (3).

[0019] Preferably, step S1 specifically further includes:

[0020] In the observation mode of push-broom strip imaging, for any two push-broom strip imaging tasks and the aggregation observation conditions are:

[0021] (4);

[0022] Among them, is the maximum continuous shooting time for single imaging.

[0023] Preferably, for the push-broom strip imaging task and , the corresponding aggregation task 's characteristic parameters include the yaw angle interval and the minimum tracking arc segment ;

[0024] Among them, The calculation method is:

[0025] (5);

[0026] The calculation method of is the same as formula (3).

[0027] Preferably, the generation of the aggregation scheme by the pre-aggregation algorithm described in step S2 specifically includes:

[0028] Step 21: Sort the task requirements according to the principle of dense area first;

[0029] Step 22: Sequentially take out the task from the task sequence , and aggregate it with the subsequent tasks in the sequence;

[0030] Step 23: Merge or deduplicate the aggregation tasks with the same single-task composition or the inclusion relationship in the task set;

[0031] Step 24: Obtain the reduced aggregation task set through the above steps, and select the aggregation tasks according to the strategy of the fewest tasks and the largest coverage. The constraint conditions of this strategy are as follows:

[0032] (6);

[0033] Among them, is used to indicate the inclusion situation of the subtask set pair for. If , let ; otherwise, let .

[0034] Preferably, step S21 specifically includes:

[0035] Step 211: Use the K-mean algorithm to cluster based on the task distribution to obtain several cluster centers ;

[0036] Step 212: For , calculate the distance between and each cluster center in , set it as , and let serve as the distance from to the cluster center;

[0037] Step 213: Sort according to the distance from the clustering center point, and regenerate the task sequence as the input for the aggregation operation.

[0038] Preferably, step S22 specifically includes:

[0039] Step 221: If is a fixed strip imaging task and satisfies formula (1), calculate the aggregated visible window and the side-sway angle interval according to formulas (2) and (3);

[0040] Step 222: If is a push-broom strip imaging task and satisfies formula (4), then calculate the aggregated minimum tracking arc segment and the side-sway angle interval according to formulas (5) and (3);

[0041] Step 223: If it can be aggregated into a new task , then use to replace , and continue to participate in the aggregation until it can no longer be aggregated.

[0042] Preferably, the dynamic adjustment of the aggregation scheme by the in-process aggregation algorithm described in step S3 specifically includes:

[0043] Step 31: Classify all tasks, and set the set of tasks that have not been satisfied or newly added as , and the set of tasks that have been satisfied and are waiting to be executed as ;

[0044] Step 32: For , calculate the number of tasks in that can be aggregated with ; for , calculate the number of tasks in that can be aggregated with ;

[0045] Step 33: Re-sort the tasks in and in ascending order according to and respectively;

[0046] Step 34: Take out the tasks from in turn, and aggregate them with the tasks in ;

[0047] Step 35: Hand to the resource scheduling system as a separate task. If it still cannot be satisfied, then add to the set of tasks that cannot be executed ;

[0048] Step 36: Remove from . If the condition is satisfied, the dynamic adjustment ends; otherwise, go to Step 32.

[0049] Preferably, Step S34 specifically includes:

[0050] Step 341: For , if meets the aggregation condition, go to Step 342; otherwise, go to Step 35;

[0051] Step 342: Conduct a constraint check on the task aggregation result. If the satellite resource allocation constraint is satisfied, generate an aggregated task and update the resource usage plan; otherwise, go to Step 35.

[0052] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0053] The present invention takes into account the imaging mode differences of earth observation payloads, sorts out the basic conditions required for the fixed strip imaging mode and the push-broom strip imaging mode respectively, and then designs different aggregation constraint condition models, thereby reducing the impact of satellite payloads on the possibility of target aggregation observation in different imaging modes; moreover, on the basis of conventional pre-event static aggregation, the present invention attempts to integrate the aggregation task into the resource scheduling process, regenerate the aggregation task according to the unmet task distribution in the original scheduling plan and the currently generated plan, and hand it over to the scheduling system for processing after evaluation, that is, realize in-event dynamic aggregation, so as to improve the sufficiency and effectiveness of the task aggregation operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 It is a schematic diagram of the effect of the aggregation observation of multiple point target observation tasks by an observation satellite;

[0055] Figure 2 It is a schematic flow chart of the pre-event aggregation algorithm and the in-event aggregation algorithm in a target aggregation method considering the imaging mode differences of earth observation payloads provided by the embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0056] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0057] Embodiments of the present invention provide a target aggregation method considering the imaging mode differences of earth observation payloads, including the following steps:

[0058] S1: Determine whether the aggregation observation conditions are met for multiple point target observations;

[0059] S2: When the aggregation observation conditions are met, provide an initial task input to the resource scheduling system and generate an aggregation plan through a pre-aggregation algorithm;

[0060] S3: Dynamically adjust the aggregation plan through an in-process aggregation algorithm according to the actual task fulfillment situation during the resource scheduling process.

[0061] According to the imaging mechanism of satellite payloads, if the observation angles of multiple point targets are within a certain range and the imaging times for them have a sufficient length of intersection (or meet the splicing conditions), aggregation observation can be considered. The existing satellite observation imaging modes include two observation modes: fixed strip imaging (mainly considering optical area array mode, SAR spotlight mode, SAR strip mode) and pushbroom strip imaging (mainly considering optical linear array mode, SAR scanning mode). The influence of satellite payloads on the possibility of target aggregation observation is different in different imaging modes. Therefore, the present invention sorts out the basic conditions required for the fixed strip imaging mode and the pushbroom strip imaging mode respectively, and then designs different aggregation constraint condition models.

[0062] In an embodiment of the present invention, step S1 specifically includes:

[0063] Let the earth observation satellite imaging requirement for point targets the t th visible window be , and be the start time and end time of the window respectively, be the payload side-sway angle corresponding to this visible window;

[0064] In the observation mode of fixed strip imaging, for any two fixed strip imaging tasks and the aggregation observation conditions are:

[0065] (1);

[0066] Among them, is the set of visible windows for ; is the minimum working time required for a single imaging; is Instantaneous Field of View (IFOV) of the payload.

[0067] Further, when the fixed-strip imaging task on and on meet the aggregation observation conditions, the aggregation task is denoted as , and its characteristic parameters include the visible window and the slew angle interval ;

[0068] Among them, The calculation method of is:

[0069] (2);

[0070] The calculation method of is:

[0071] (3).

[0072] And, step S1 specifically further includes:

[0073] In the push-broom strip imaging observation mode, for any two push-broom strip imaging tasks and the aggregation observation conditions are:

[0074] (4);

[0075] Among them, is The maximum continuous shooting time for a single imaging.

[0076] Further, for the push-broom strip imaging tasks and , the corresponding aggregation task The characteristic parameters include the slew angle interval and the minimum tracking arc segment ;

[0077] Among them, The calculation method of is:

[0078] (5);

[0079] The calculation method of is the same as formula (3).

[0080] For the problem of task aggregation, a common solution is to perform pre-integration, that is, before starting resource scheduling, first merge the point targets according to certain rules to form several point target sets (aggregated tasks), and each set as a whole can participate in satellite resource allocation as a separate task. Pre-integration does not need to consider the actual use of satellite resources. It belongs to pre-static aggregation. The implementation method is relatively simple and suitable for rapid analysis of task distribution status and preliminary formulation of aggregation schemes. On this basis, the present invention attempts to integrate task integration into the resource scheduling process, that is, to achieve dynamic aggregation during the process, so as to improve the adequacy and effectiveness of task aggregation operations.

[0081] Figure 2 The flowchart of the pre-aggregation algorithm and the in-process aggregation algorithm in a target aggregation method considering the differences in imaging modes of earth observation payloads provided by an embodiment of the present invention is shown. Figure 2 As shown, the overall aggregation process formed by the aggregation scheme of the present invention is as follows: after the demand is clear, first generate the aggregation task according to the basic constraints (the subtasks can overlap). According to certain rules, evaluate the overall importance or priority of the aggregation task. Submit the aggregation task to the satellite resource scheduling system to participate in centralized selection and resource allocation. According to the original task situation that is not met and the currently generated plan, regenerate the aggregation task and hand it over to the resource scheduling system for processing after evaluation. Repeat the above process until the plan output conditions are met.

[0082] In one embodiment of the present invention, the generation of the aggregation scheme by the pre-aggregation algorithm in step S2 specifically includes:

[0083] Step 21: Sort the task requirements according to the principle of giving priority to densely populated areas;

[0084] Step 22: From the Task Sequence Take out the tasks one by one , aggregated with subsequent tasks in the sequence;

[0085] Step 23: Merge or remove duplicates of aggregated tasks that have the same individual task composition or have a containment relationship in the task set;

[0086] Step 24: Obtain the simplified aggregated task set through the above steps, and select the aggregated tasks according to the strategy of minimum tasks and maximum coverage. The constraints of this strategy are as follows:

[0087] (6);

[0088] in, Used for marking Subtask set pair Including situations, if ,make ; Otherwise, let .

[0089] Furthermore, step S21 specifically includes:

[0090] Step 211: Use the K-mean algorithm to perform clustering based on the task distribution to obtain several cluster centers ;

[0091] Step 212: For , calculate and the distances from each cluster center in , let be the distance from

[0092] Step 213: Sort according to the distances from the cluster centers and regenerate the task sequence as the input for the aggregation operation.

[0093] Step S22 specifically includes:

[0094] Step 221: If is a fixed-strip imaging task and satisfies Equation (1), calculate the aggregated visible window and the side-sway angle range according to Equations (2) and (3);

[0095] Step 222: If is a push-broom strip imaging task and satisfies Equation (4), then calculate the aggregated minimum tracking arc segment and the side-sway angle range according to Equation (5) and Equation (3);

[0096] Step 223: If it can be aggregated into a new task , then use to replace , and continue to participate in the aggregation until no further aggregation is possible.

[0097] The task pre-aggregation in the present invention mainly provides the initial task input for the resource scheduling system. During the aggregation operation, the actual resource allocation situation is usually not considered. While the task in-process aggregation mainly dynamically adjusts the aggregation scheme according to the actual task satisfaction situation during the resource scheduling process, and assists the resource scheduling system in generating the final task execution plan. In an embodiment of the present invention, the dynamic adjustment of the aggregation scheme through the in-process aggregation algorithm described in step S3 specifically includes:

[0098] Step 31: Classify all tasks, and let the set of tasks that have not been satisfied or newly added be , and the set of tasks that have been satisfied and are pending execution be ;

[0099] Step 32: For , calculate The number of tasks that can be aggregated with ; for ; calculate the number of tasks that can be aggregated with ; ; ;

[0100] Step 33: Re - sort the tasks in and in ascending order according to and respectively;

[0101] Step 34: Take out the tasks from in turn and aggregate them with the tasks in ;

[0102] Step 35: Take as a separate task and hand it over to the resource scheduling system. If it still cannot be satisfied, then add to the set of tasks that cannot be executed ;

[0103] Step 36: Remove from . If the condition is satisfied, the dynamic adjustment ends; otherwise, go to Step 32.

[0104] Furthermore, Step S34 specifically includes:

[0105] Step 341: For , if meets the aggregation condition, then go to Step 342; otherwise, go to Step 35;

[0106] Step 342: Conduct constraint checking on the task aggregation result. If the satellite resource allocation constraint is satisfied, generate an aggregated task and update the resource usage plan; otherwise, go to Step 35.

[0107] Although the embodiments of the present invention have been shown and described, for those of ordinary skill in the art, it can be understood that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A target aggregation method considering the differences in imaging modes of earth observation payloads, characterized in that: The following steps are involved: S1: Determine whether multiple point target observations meet the conditions for aggregate observation; S2: Provide the resource scheduling system with the initial task input when the aggregation observation conditions are met, and generate an aggregation plan through the prior aggregation algorithm; S3: According to the actual task satisfaction during resource scheduling, the aggregation scheme is dynamically adjusted through the in-process aggregation algorithm; Step S1 specifically includes: Earth observation satellite Point target imaging requirements No. t The visible window is , and are the start and end times of the window, respectively. is the load side swing angle corresponding to the visible window; In the observation mode of fixed strip imaging, for any two fixed strip imaging tasks and The aggregation observation conditions are: （1）； in, for right The visible window collection of for Minimum working time required for a single imaging session; for The instantaneous field of view of the load; When fixed strip imaging tasks exist Shanghe exist When the aggregation observation conditions are met, the aggregation task is recorded as , whose characteristic parameters include the visible window and roll angle range ; in, The calculation method is: （2）； The calculation method is: （3）； Step S1 specifically also includes: In the push-scan strip imaging observation mode, for any two push-scan strip imaging tasks and The aggregation observation conditions are: （4）； in, for Maximum continuous shooting time for a single imaging; For push-scan swath imaging tasks and , the corresponding aggregation task The characteristic parameters include the side swing angle range and the minimum tracking arc ; in, The calculation method is: （5）； The calculation method of is the same as formula (3); The dynamic adjustment of the aggregation scheme by the in-progress aggregation algorithm described in step S3 specifically includes: Step 31: Classify all tasks and set the set of tasks that have not been met or newly added as , the set of tasks that have been satisfied and are to be executed is ; Step 32: For ,calculate Zhongke The number of tasks to aggregate ;for ,calculate Zhongke The number of tasks to aggregate ; Step 33: and The tasks are divided into and Reorder in ascending order; Step 34: Take out the tasks in turn, and Aggregate the tasks in Step 35: As a separate task to the resource scheduling system, if it still cannot be satisfied, then Add to the collection of unavailable tasks ; Step 36: from Remove if conditions are met , dynamic adjustment ends; otherwise, go to step 32.

2. The target aggregation method considering the differences in imaging modes of earth observation payloads according to claim 1 is characterized in that: The step S2 of generating an aggregation scheme by using a pre-aggregation algorithm specifically includes: Step 21: Sort the task requirements according to the principle of giving priority to densely populated areas; Step 22: From the Task Sequence Take out the tasks one by one , aggregated with subsequent tasks in the sequence; Step 23: Merge or remove duplicates of aggregated tasks that have the same individual task composition or have a containment relationship in the task set; Step 24: Obtain the simplified aggregated task set through the above steps, and select the aggregated tasks according to the strategy of minimum tasks and maximum coverage. The constraints of this strategy are as follows: （6）； in, Used for marking Subtask set pair If ,make Otherwise, let .

3. The target aggregation method considering the differences in imaging modes of earth observation payloads according to claim 2 is characterized in that: Step S21 specifically includes: Step 211: Use the K-mean algorithm to perform clustering based on task distribution and obtain several cluster center points ; Step 212: For ,calculate and The distance between each cluster center point in is set to ,make As Distance from the cluster center; Step 213: Sort by distance from the cluster center and regenerate the task sequence as input for the aggregation operation.

4. The target aggregation method considering the differences in imaging modes of earth observation payloads according to claim 3 is characterized in that: Step S22 specifically includes: Step 221: If For a fixed strip imaging task, and satisfying equation (1), the aggregated visible window and side swing angle interval are calculated according to equations (2) and (3); Step 222: If For push-scan strip imaging tasks, and satisfying equation (4), the minimum tracking arc segment and side swing angle interval after aggregation are calculated according to equations (5) and (3); Step 223: If it can be aggregated into a new task , then Alternative , continue to participate in aggregation until aggregation is no longer possible.

5. The target aggregation method considering the differences in imaging modes of earth observation payloads according to claim 1, characterized in that: Step S34 specifically includes: Step 341: For ,like If the aggregation condition is met, proceed to step 342; otherwise, proceed to step 35; Step 342: Perform constraint check on the task aggregation result. If the satellite resource allocation constraint is met, generate an aggregation task and update the resource usage plan; otherwise, go to step 35.

Images