CN111178419B - Agile remote sensing satellite multi-target task planning method based on task clustering - Google Patents
Agile remote sensing satellite multi-target task planning method based on task clustering Download PDFInfo
- Publication number
- CN111178419B CN111178419B CN201911349972.0A CN201911349972A CN111178419B CN 111178419 B CN111178419 B CN 111178419B CN 201911349972 A CN201911349972 A CN 201911349972A CN 111178419 B CN111178419 B CN 111178419B
- Authority
- CN
- China
- Prior art keywords
- cluster
- task
- target
- point
- observation
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
- G06Q10/047—Optimisation of routes or paths, e.g. travelling salesman problem
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- Human Resources & Organizations (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Economics (AREA)
- Evolutionary Biology (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Strategic Management (AREA)
- Evolutionary Computation (AREA)
- Game Theory and Decision Science (AREA)
- Development Economics (AREA)
- Entrepreneurship & Innovation (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
- Image Processing (AREA)
Abstract
The invention relates to a task clustering-based agile remote sensing satellite multi-target task planning method, which comprises the following steps: (1) Converting an observation target into discrete point sets, clustering and grouping the target sets to be observed by a boundary continuation method under a greedy strategy, determining each cluster group as an inner layer task set, and determining the combination of all inner layer task sets as an outer layer task set; (2) A local tabu algorithm is adopted for the inner layer task set, and an inner layer task observation sequence is planned; (3) And (3) adopting a global tabu algorithm for the outer layer task set to plan an optimal overall observation path. Compared with the prior art, the method and the device effectively reduce the complexity of task planning, avoid the situation that planning falls into local optimum, improve the probability of obtaining the optimum solution, reduce the problem solving neighborhood and greatly reduce the calculation time of an algorithm.
Description
Technical Field
The invention relates to the technical field of agile remote sensing satellites, in particular to a task clustering-based agile remote sensing satellite multi-target task planning method.
Background
The advanced remote sensing technology of the new system is explored, the high-score, high-efficiency and high-value earth observation satellite system with international leading level is built, the timeliness and the accuracy of earth observation information acquisition are improved, and the method is significant. The agile remote sensing satellite multi-target task planning technology and the planning system based on task clustering are researched, and a certain theoretical result and technical reserve can be provided for autonomous task planning and operation of a new generation agile satellite in China.
Agile satellites are used as a new remote sensing satellite, and have been rapidly applied and developed in recent years due to the advantages of various imaging modes, high camera resolution, wide gesture maneuvering range, high speed, high precision and the like. The observation path of the satellite refers to a shooting route when the camera moves along the specified direction along with the satellite, and the design of the observation path is an important ring of satellite mission planning. Compared with the traditional imaging mode that the remote sensing satellite maneuvers and shoots at first, the agile satellite can shoot at the same time, namely the satellite performs push-broom imaging on the ground target while maneuvering in the gesture. The new imaging modes present new challenges for satellite observation path design. The observation direction of a camera of a traditional remote sensing satellite is fixed in the imaging process, and the observation path of the satellite is a plurality of parallel strips parallel to the undersea point; and the agile satellite can observe along any direction due to the special technology of 'imaging in motion'. Therefore, in the task planning process, the design scheme of the observation path for dividing parallel strips in the traditional method needs to be changed, and a plurality of observation paths consisting of broken lines or curves are designed according to satellite imaging constraint conditions and the priority level of an observation target.
In the whole agile satellite observation path design process, specific path design schemes need to be formulated according to the characteristics of observation targets in a classified manner. Currently, observation targets are generally classified into point targets and area targets according to state characteristics. In the existing planning scheme, whether the point target or the area target is oriented, a single point target is used as a scheduling object by a researcher, the number of planning candidate solutions increases in geometric magnitude with the increase of the number of tasks, and a planning system bears a large amount of calculation burden under the condition of limited calculation capacity. In practice, as the on-board sensor has a certain angle of view, an observation strip with a certain width is generated on the ground when the on-board sensor passes through the ground, if the actual observation strip is taken as a basis to be combined with an imaging constraint condition, the observation path of the satellite is planned according to the target distribution condition according to local conditions, and the observation efficiency of the agile remote sensing satellite can be greatly improved.
Disclosure of Invention
The invention aims to overcome the defects of the prior art and provide an agile remote sensing satellite multi-target task planning method based on task clustering.
The aim of the invention can be achieved by the following technical scheme:
a task clustering-based agile remote sensing satellite multi-target task planning method comprises the following steps:
(1) Converting an observation target into discrete point sets, clustering and grouping the target sets to be observed by a boundary continuation method under a greedy strategy, determining each cluster group as an inner layer task set, and determining the combination of all inner layer task sets as an outer layer task set;
(2) A local tabu algorithm is adopted for the inner layer task set, and an inner layer task observation sequence is planned;
(3) And (3) adopting a global tabu algorithm for the outer layer task set to plan an optimal overall observation path.
The step (1) comprises the following steps:
(11) Constructing an original Cluster set Cluster x : with any one of the target points post in the discrete point set x Calculating all points post in the discrete point set as the center point i And the target point post x Is to assign the points for which Dist (x, i) is less than the threshold Mdist of the extension distance to the original Cluster set Cluster x The method comprises the following steps:
Cluster x ＝{Pot i |1≤i≤N,Dist(x,i)＜Mdist}，
wherein N is the total number of target points in the discrete point set;
(12) Building an extended Cluster set Cluster x_exp : computing an original Cluster x In (c) and the set boundary peripheral target point post j Extending points with a distance less than the extended distance threshold Mdist to the original Cluster set Cluster x In forming an extended Cluster set Cluster x_exp The method comprises the following steps:
Cluster x_exp ＝{Pot j |1≤i≤M,Pot k ∈Cluster x ,Dist(k,j)＜Mdist}，
wherein Dist (k, j) is the original Cluster set Cluster x Target point of (2) post k And the boundary peripheral target point post of the original cluster set j M is the total number of peripheral target points of the boundary of the original cluster set;
(13) And (3) taking the expanded clustering set as a clustering group, removing clustered target points from the discrete point set, and repeating the steps (11) - (12) to finish clustering of all the target points.
In the clustering process, a single target point which cannot be clustered is taken as an independent inner layer Task set, namely, the finally determined outer layer Task set is the Task:
Task＝{Pot 1 ,Pot 2 ,…,Pot RE ,Cluster x_exp1 ,Cluster x_exp2 ,…,Cluster x_expn }，
wherein, the post 1 For the 1 st target point which cannot be clustered, post 2 For the 2 nd target point which cannot be clustered, post RE For the RE target points which cannot be clustered, RE is the total number of target points which cannot be clustered, cluster x_exp1 Cluster for group 1 comprising at least 2 target points x_exp2 Cluster for the 2 nd Cluster group comprising at least 2 target points x_expn For the nth cluster group comprising at least 2 target points, n represents the total number of cluster groups comprising at least 2 target points.
The step (2) comprises the following steps:
cluster for any one inner layer task set x_exp Setting total Nsub target points in the set, and selecting two target points with the largest relative distance under the longitude and latitude coordinate system as the starting points post of the target observation sequence start And end point post end The other (Nsub-2) target points in the set are sequentially sequenced as intermediate target points to form an initial optimization sequence InSub_Cluster x_exp ＝{Pot start ,Pot 1 ,Pot 2 ,…,Pot Nsub-2 ,Pot end The method comprises the steps of (a) re-ordering an initial optimization sequence to generate an exchange sequence by adopting an inner layer tabu search algorithm to exchange sequences in a set of points in pairs aiming at the initial optimization sequence, calculating the benefit of the exchange sequence, and selecting the exchange sequence with the highest benefit value as an optimization sequence sub_Cluster x_exp Optimizing sequence sub_Cluster x_exp The ordering of the middle target points is an inner layer task set Cluster x_exp And observing the sequence of the middle-inner layer task.
The benefit of the exchange sequence is obtained by:
F＝∑ i∈Nallowed w i ·x i +1/∑ i,j∈Nallowed Shift(i,j)，
wherein F represents benefit, w i For task weight, x i For the observation coefficient of the ith target point in the exchange sequence, x is when the target point is observable i Get 1, otherwise x i Taking 0, shift (i, j) to represent camera attitude adjustment time between an ith target point and a jth target point in the exchange sequence, and Nallowed is the exchange sequence target point set.
The step (3) is specifically as follows:
(31) Determining an initial solution of an optimal overall observation path;
(32) Judging the observation constraint conditions according to the set observation target time window, and eliminating points which cannot be observed in the current overall observation path;
(33) Obtaining the benefit of the current overall observation path;
(34) Taking each cluster group in the current overall observation path as a point, adopting a global tabu search algorithm to exchange the sequence of the midpoints of the current overall observation path pairwise, and repeatedly executing (32) - (33) until the maximum iteration times are reached;
(35) And selecting the overall observation path with the biggest profit as the final solution of the optimal overall observation path.
The step (31) is specifically as follows:
and taking each cluster group as a point, sequentially sequencing the benefits of each cluster group from large to small to form a set InW _task, starting from the first position to the last position of the set InW _task, exchanging element positions two by two to form a new set WCH_task, calculating the benefits of the set, and selecting the WCH_task set with the highest benefits as an initial solution of the optimal overall observation path.
The step (32) is specifically as follows:
selecting the starting time of a first target point as the beginning of a time line, adding the observation time l on the time line, calculating the posture transfer time of a satellite for observing adjacent target points, and adding the posture transfer time into the time line to obtain the time t after posture transfer 1 Judgment of t 1 Relation to start time st and et of next target point observation time window, if t 1 <st, start time t of next target point 2 =st, if st<t 1 <et, t 1 For the start time of the next target point, if t 1 >et, the next target point is determined to be an unobservable point, which is culled.
The cluster group benefits or the aggregate benefits or the benefits of the current overall observation path are obtained by the following modes:
F＝∑ i∈X w i ·x i +1/∑ i,j∈X Shift(i,j)，
wherein F represents benefit, w i For task weight, x i As the observation coefficient of the ith target point, x is when the target point is observable i Get 1, otherwise x i Taking 0, and shifting (i, j) to represent camera attitude adjustment time between the ith target point and the jth target point, wherein X is the number of target points in a corresponding cluster group or set or current whole observation path.
Compared with the prior art, the invention has the following advantages:
(1) According to the method, a boundary continuation method under a greedy strategy is adopted, a large-scale target is clustered into different task sets according to the observation view field of the satellite camera, and the complexity of task planning is effectively reduced;
(2) According to the invention, detour search is effectively avoided through a local neighborhood search mechanism and corresponding tabu criteria in the inner and outer layer tabu algorithm, and specific tabu objects are released through the scoffy criteria, so that diversity of solution sets is ensured to a certain extent, and global optimization is finally realized;
(3) According to the method, the target set is classified and processed in a hierarchical planning mode, the neighborhood of the feasible solution of the problem is effectively reduced, and the calculation time of an algorithm is greatly reduced.
Drawings
FIG. 1 is a general flow chart of a task clustering-based agile remote sensing satellite multi-target task planning method of the invention;
FIG. 2 is a schematic diagram of the results of obtaining a cluster panel according to the present invention;
FIG. 3 is a schematic diagram of the result of the observation sequence of the inner layer task based on the tabu algorithm in the embodiment of the present invention;
fig. 4 is a schematic diagram of the result of the optimal overall observation path in the embodiment of the present invention.
Detailed Description
The invention will now be described in detail with reference to the drawings and specific examples. Note that the following description of the embodiments is merely an example, and the present invention is not intended to be limited to the applications and uses thereof, and is not intended to be limited to the following embodiments.
Examples
Taking the polar orbit remote sensing satellite shown in table 1 as an example, the 9 point-shaped target observation task plans in table 2 are taken as an implementation procedure of the invention, and an embodiment of the scheme of the invention is described in detail with reference to the accompanying drawings.
TABLE 1 number of polar orbit remote sensing satellites orbits
Table 2 nine-point target information
As shown in FIG. 1, the method for planning the multiple target tasks of the agile remote sensing satellite based on task clustering comprises three steps.
As shown in fig. 2, step 1 specifically includes:
(11) Constructing an original Cluster set Cluster x : with any one of the target points post in the discrete point set x Calculating a discrete value for the center pointAll points post in a point set i And the target point post x Is to assign the points for which Dist (x, i) is less than the threshold Mdist of the extension distance to the original Cluster set Cluster x The method comprises the following steps:
Cluster x ＝{Pot i |1≤i≤N,Dist(x,i)＜Mdist}，
wherein N is the total number of target points in the discrete point set;
(12) Building an extended Cluster set Cluster x_exp : computing an original Cluster x In (c) and the set boundary peripheral target point post j Extending points with a distance less than the extended distance threshold Mdist to the original Cluster set Cluster x In forming an extended Cluster set Cluster x_exp The method comprises the following steps:
Cluster x_exp ＝{Pot j |1≤i≤M,Pot k ∈Cluster x ,Dist(k,j)＜Mdist}，
wherein Dist (k, j) is the original Cluster set Cluster x Target point of (2) post k And the boundary peripheral target point post of the original cluster set j M is the total number of peripheral target points of the boundary of the original cluster set;
(13) Taking the expanded Cluster set as a Cluster group, removing clustered target points from the discrete point set and repeating the steps (11) - (12) to finish the clustering of all the target points, namely obtaining an expanded Cluster set Cluster in FIG. 1 x_exp Then, selecting a target point post z Repeating steps (11) - (12) for the center point.
In the clustering process, a single target point which cannot be clustered is taken as an independent inner layer Task set, namely, the finally determined outer layer Task set is the Task:
Task＝{Pot 1 ,Pot 2 ,…,Pot RE ,Cluster x_exp1 ,Cluster x_exp2 ,…,Cluster x_expn }，
wherein, the post 1 For the 1 st target point which cannot be clustered, post 2 For the 2 nd target point which cannot be clustered, post RE RE is the RE target point which cannot be clusteredTotal number of target points that cannot be clustered, cluster x_exp1 Cluster for group 1 comprising at least 2 target points x_exp2 Cluster for the 2 nd Cluster group comprising at least 2 target points x_expn For the nth cluster group comprising at least 2 target points, n represents the total number of cluster groups comprising at least 2 target points.
And 2, adopting a local tabu algorithm to the inner layer task set to plan an inner layer task observation sequence.
The step 2 is specifically as follows:
cluster for any one inner layer task set x_exp Setting total Nsub target points in the set, and selecting two target points with the largest relative distance under the longitude and latitude coordinate system as the starting points post of the target observation sequence start And end point post end The other (Nsub-2) target points in the set are sequentially sequenced as intermediate target points to form an initial optimization sequence InSub_Cluster x_exp ＝{Pot start ,Pot 1 ,Pot 2 ,…,Pot Nsub-2 ,Pot end The method comprises the steps of (a) re-ordering an initial optimization sequence to generate an exchange sequence by adopting an inner layer tabu search algorithm to exchange sequences in a set of points in pairs aiming at the initial optimization sequence, calculating the benefit of the exchange sequence, and selecting the exchange sequence with the highest benefit value as an optimization sequence sub_Cluster x_exp Optimizing sequence sub_Cluster x_exp The ordering of the middle target points is an inner layer task set Cluster x_exp And observing the sequence of the middle-inner layer task.
The benefit of the exchange sequence is obtained by:
F＝∑ i∈Nallowed w i ·x i +1/∑ i,j∈Nallowed Shift(i,j)，
wherein F represents benefit, w i For task weight, x i For the observation coefficient of the ith target point in the exchange sequence, x is when the target point is observable i Get 1, otherwise x i Taking 0, shift (i, j) to represent camera attitude adjustment time between an ith target point and a jth target point in the exchange sequence, and Nallowed is the exchange sequence target point set.
In the above stepsIn step 2, the length L of the tabu table is set ncb The length of the tabu table does not exceed the number of candidate sequences. In general, the length of the tabu table is smaller, the algorithm is easy to be locally optimized, and the length of the tabu table is larger, so that the algorithm is difficult to converge. A for all candidate sequences of the run ij Switching, sorting all sequences according to an evaluation function, taking the front L ncb The individual sequences are placed in a tabu list. Setting initial tabu number N ncb . If A is randomly generated in the next round ij In the taboo table, the number of taboos is reduced by 1 and a random A is regenerated ij 。
The candidate solutions are changed according to the domain functions, a plurality of domain solutions are generated, the solution with the optimal state is selected from the generated domain solutions to serve as a new candidate solution, the candidate solution is judged according to the scofflaw, if the scofflaw is met, the candidate solution is used for replacing the current solution, the tabu object in front of the tabu list is deleted, the object corresponding to the current optimal state is added to serve as a new tabu object, the current optimal state is replaced, various objects corresponding to the candidate solution are analyzed, the tabu attribute is judged, the object in the candidate solution set is judged, the object which is not the tabu object in the tabu list is used as the new current solution, the optimal state corresponding to the object is used as the new current solution, and the original tabu object is replaced by the new tabu object. When the iteration number of the algorithm is greater than a certain preset value, the end algorithm outputs an observation sequence inside the cluster, as shown in fig. 3, where (a) in fig. 3 is the distribution of all target points, (B) in fig. 3 is a cluster group obtained by clustering, the embodiment is clustered into 3 cluster groups, and (c) in fig. 3 is an inner task observation sequence obtained by performing inner task planning on the cluster group B in (B) in fig. 3.
And 3, planning an optimal overall observation path by adopting a global tabu algorithm to the external task set. The method comprises the steps of adopting an externally nested tabu algorithm to intensively plan clustered sets and scattered point targets, and finally generating a global observation path. This step involves the following parts:
1) Initial solution generation
The clustering set has two observation directions, and the optimal observation direction needs to be screened out according to the optimization index. High quality initial solutions are also important factors to ensure the effectiveness of the tabu algorithm.
2) Evaluation function
Because the satellite single transit observation capability is limited, the single transit cannot meet all user observation requirements. In order to ensure that as many observations as possible are made for important targets within a limited time. The invention takes the overall income of the task as a primary optimization index, and the observation time of the task as a secondary optimization index, so that the evaluation function is as follows:
maximize sigma i∈X w i ·x i
Minimizing sigma i,j∈X Shift(i,j)
Maximizing f= Σ i∈X w i ·x i +1/∑ i,j∈X Shift(i,j)
F represents the benefit, w i For task weight, x i As the observation coefficient of the ith target point, x is when the target point is observable i Get 1, otherwise x i Taking 0, shift (i, j) to represent the camera attitude adjustment time between the ith target point and the jth target point, and X is the number of target points in the observation path.
3) Establishing a clustered target observation time window
Let TW i ＝[st i ,et i ]For satellite to task i scout time window, st i To start time, et i For the end time. After clustering of the plurality of tasks, a time window needs to be determined for the new clustered task. The method of the invention superimposes the time windows of a plurality of tasks to be used as the time windows of the clustering tasks. Assuming TW is present j ＝[st j ,et j ]Overlapping with the task, the time window of the clustering task is set as [ st ] i ,et j ]。
4) Determination of observation constraint conditions
It is determined whether the nodes in the solution are observable. Selecting the starting time of a first node as the beginning of a time line, and adding the observation time l on the time line; calculating the attitude transition time Shift (i, j) of the satellite observation adjacent node, andthe gesture transition time is added to the timeline. Judgment of t 1 Relation to start time st and et of next target point observation time window, if t 1 <st, start time t of next target point 2 =st, if st<t 1 <et, t 1 For the start time of the next target point, if t 1 >et, then the next target point is determined to be an unobservable point.
In summary, step 3 specifically includes:
(31) Determining an initial solution of an optimal overall observation path;
(32) Judging the observation constraint conditions according to the set observation target time window, and eliminating points which cannot be observed in the current overall observation path;
(33) Obtaining the benefit of the current overall observation path;
(34) Taking each cluster group in the current overall observation path as a point, adopting a global tabu search algorithm to exchange the sequence of the midpoints of the current overall observation path pairwise, and repeatedly executing (32) - (33) until the maximum iteration times are reached;
(35) And selecting the overall observation path with the biggest profit as the final solution of the optimal overall observation path.
The step (31) is specifically as follows:
and taking each cluster group as a point, sequentially sequencing the benefits of each cluster group from large to small to form a set InW _task, starting from the first position to the last position of the set InW _task, exchanging element positions two by two to form a new set WCH_task, calculating the benefits of the set, and selecting the WCH_task set with the highest benefits as an initial solution of the optimal overall observation path.
The step (32) is specifically as follows:
selecting the starting time of a first target point as the beginning of a time line, adding the observation time l on the time line, calculating the posture transfer time of a satellite for observing adjacent target points, and adding the posture transfer time into the time line to obtain the time t after posture transfer 1 Judgment of t 1 Relation to start time st and et of next target point observation time window, if t 1 <st, start of next target pointTime t 2 =st, if st<t 1 <et, t 1 For the start time of the next target point, if t 1 >et, the next target point is determined to be an unobservable point, which is culled.
Clustering the group benefits or the aggregate benefits or the benefits of the current overall observation path is achieved by:
F＝∑ i∈X w i ·x i +1/∑ i,j∈X Shift(i,j)，
wherein F represents benefit, w i For task weight, x i As the observation coefficient of the ith target point, x is when the target point is observable i Get 1, otherwise x i Taking 0, and shifting (i, j) to represent camera attitude adjustment time between the ith target point and the jth target point, wherein X is the number of target points in a corresponding cluster group or set or current whole observation path.
The final solution of the optimal overall observation path in this embodiment is shown in fig. 4.
According to the invention, a large-scale discrete target is divided into different task sets by a boundary continuation method under a greedy strategy, so that the complexity of task planning is effectively reduced; the optimal task sequences and optimal observation paths of different levels are calculated through the inner-outer tabu algorithm, so that the situation that planning falls into local optimal is effectively avoided, and the probability of obtaining an optimal solution is improved; through a hierarchical planning mode, the problem solving neighborhood is effectively reduced, and the calculation time of an algorithm is greatly reduced.
The above embodiments are merely examples, and do not limit the scope of the present invention. These embodiments may be implemented in various other ways, and various omissions, substitutions, and changes may be made without departing from the scope of the technical idea of the present invention.
Claims (7)
1. The agile remote sensing satellite multi-target task planning method based on task clustering is characterized by comprising the following steps of:
(1) Converting an observation target into discrete point sets, clustering and grouping the target sets to be observed by a boundary continuation method under a greedy strategy, determining each cluster group as an inner layer task set, and determining the combination of all inner layer task sets as an outer layer task set;
(2) A local tabu algorithm is adopted for the inner layer task set, and an inner layer task observation sequence is planned;
(3) A global tabu algorithm is adopted for the outer layer task set, and an optimal overall observation path is planned;
the step (2) comprises the following steps:
cluster for any one inner layer task set x_exp Setting total Nsub target points in the set, and selecting two target points with the largest relative distance under the longitude and latitude coordinate system as the starting points post of the target observation sequence start And end point post end The other (Nsub-2) target points in the set are sequentially sequenced as intermediate target points to form an initial optimization sequence InSub_Cluster x_exp ＝{Pot start ,Pot 1 ,Pot 2 ,…,Pot Nsub-2 ,Pot end The method comprises the steps of (a) re-ordering an initial optimization sequence to generate an exchange sequence by adopting an inner layer tabu search algorithm to exchange sequences in a set of points in pairs aiming at the initial optimization sequence, calculating the benefit of the exchange sequence, and selecting the exchange sequence with the highest benefit value as an optimization sequence sub_Cluster x_exp Optimizing sequence sub_Cluster x_exp The ordering of the middle target points is an inner layer task set Cluster x_exp A middle-inner layer task observation sequence;
the benefit of the exchange sequence is obtained by:
F＝∑ i∈Nallowed w i ·x i +1/∑ i,j∈Nallowed Shift(i,j)，
wherein F represents benefit, w i For task weight, x i For the observation coefficient of the ith target point in the exchange sequence, x is when the target point is observable i Get 1, otherwise x i Taking 0, shift (i, j) to represent camera attitude adjustment time between an ith target point and a jth target point in the exchange sequence, and Nallowed is the exchange sequence target point set.
2. The method for planning multiple tasks of an agile remote sensing satellite based on task clustering according to claim 1, wherein the step (1) is specifically:
(11) Constructing an original Cluster set Cluster x : with any one of the target points post in the discrete point set x Calculating all points post in the discrete point set as the center point i And the target point post x Is to assign the points for which Dist (x, i) is less than the threshold Mdist of the extension distance to the original Cluster set Cluster x The method comprises the following steps:
Cluster x ＝{Pot i |1≤i≤N,Dist(x,i)＜Mdist}，
wherein N is the total number of target points in the discrete point set;
(12) Building an extended Cluster set Cluster x_exp : computing an original Cluster x In (c) and the set boundary peripheral target point post j Extending points with a distance less than the extended distance threshold Mdist to the original Cluster set Cluster x In forming an extended Cluster set Cluster x_exp The method comprises the following steps:
Cluster x_exp ＝{Pot j |1≤i≤M,Pot k ∈Cluster x ,Dist(k,j)＜Mdist}，
wherein Dist (k, j) is the original Cluster set Cluster x Target point of (2) post k And the boundary peripheral target point post of the original cluster set j M is the total number of peripheral target points of the boundary of the original cluster set;
(13) And (3) taking the expanded clustering set as a clustering group, removing clustered target points from the discrete point set, and repeating the steps (11) - (12) to finish clustering of all the target points.
3. The Task clustering-based agile remote sensing satellite multi-target Task planning method according to claim 2, wherein in the clustering process, a single target point which cannot be clustered is used as an independent inner Task set, namely, a final determined outer Task set is Task:
Task＝{Pot 1 ,Pot 2 ,…,Pot RE ,Cluster x_exp1 ,Cluster x_exp2 ,…,Cluster x_expn }，
wherein, the post 1 For the 1 st target point which cannot be clustered, post 2 For the 2 nd target point which cannot be clustered, post RE For the RE target points which cannot be clustered, RE is the total number of target points which cannot be clustered, cluster x_exp1 Cluster for group 1 comprising at least 2 target points x_exp2 Cluster for the 2 nd Cluster group comprising at least 2 target points x_expn For the nth cluster group comprising at least 2 target points, n represents the total number of cluster groups comprising at least 2 target points.
4. The method for planning multiple tasks of an agile remote sensing satellite based on task clustering according to claim 1, wherein the step (3) is specifically:
(31) Determining an initial solution of an optimal overall observation path;
(32) Judging the observation constraint conditions according to the set observation target time window, and eliminating points which cannot be observed in the current overall observation path;
(33) Obtaining the benefit of the current overall observation path;
(34) Taking each cluster group in the current overall observation path as a point, adopting a global tabu search algorithm to exchange the sequence of the midpoints of the current overall observation path pairwise, and repeatedly executing (32) - (33) until the maximum iteration times are reached;
(35) And selecting the overall observation path with the biggest profit as the final solution of the optimal overall observation path.
5. The method for planning multiple tasks of an agile remote sensing satellite based on task clustering according to claim 4, wherein the step (31) is specifically:
and taking each cluster group as a point, sequentially sequencing the benefits of each cluster group from large to small to form a set InW _task, starting from the first position to the last position of the set InW _task, exchanging element positions two by two to form a new set WCH_task, calculating the benefits of the set, and selecting the WCH_task set with the highest benefits as an initial solution of the optimal overall observation path.
6. The task clustering-based agile remote sensing satellite multi-objective task planning method according to claim 4, wherein the step (32) is specifically:
selecting the starting time of a first target point as the beginning of a time line, adding the observation time l on the time line, calculating the posture transfer time of a satellite for observing adjacent target points, and adding the posture transfer time into the time line to obtain the time t after posture transfer 1 Judgment of t 1 Relation to start time st and et of next target point observation time window, if t 1 <st, start time t of next target point 2 =st, if st<t 1 <et, t 1 For the start time of the next target point, if t 1 >et, the next target point is determined to be an unobservable point, which is culled.
7. The task clustering-based agile remote sensing satellite multi-objective task planning method according to claim 5, wherein the cluster group benefits or the aggregate benefits or the benefits of the current overall observation path are obtained by the following ways:
F＝∑ i∈X w i ·x i +1/∑ i,j∈X Shift(i,j)，
wherein F represents benefit, w i For task weight, x i As the observation coefficient of the ith target point, x is when the target point is observable i Get 1, otherwise x i Taking 0, and shifting (i, j) to represent camera attitude adjustment time between the ith target point and the jth target point, wherein X is the number of target points in a corresponding cluster group or set or current whole observation path.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911349972.0A CN111178419B (en) | 2019-12-24 | 2019-12-24 | Agile remote sensing satellite multi-target task planning method based on task clustering |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911349972.0A CN111178419B (en) | 2019-12-24 | 2019-12-24 | Agile remote sensing satellite multi-target task planning method based on task clustering |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111178419A CN111178419A (en) | 2020-05-19 |
| CN111178419B true CN111178419B (en) | 2023-05-19 |
Family
ID=70652134
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911349972.0A Active CN111178419B (en) | 2019-12-24 | 2019-12-24 | Agile remote sensing satellite multi-target task planning method based on task clustering |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111178419B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112149911B (en) * | 2020-09-30 | 2024-04-02 | 武汉大学 | Ultra-agile satellite same-orbit multipoint target in-motion imaging task planning method |
| CN114035913A (en) * | 2021-10-18 | 2022-02-11 | 南京航空航天大学 | Satellite multi-target repeated observation task planning method and terminal |
| CN115202846B (en) * | 2022-07-26 | 2023-04-07 | 北京微纳星空科技有限公司 | Method and device for determining remote control instruction of in-orbit satellite |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107239661A (en) * | 2017-06-05 | 2017-10-10 | 中国电子科技集团公司第五十四研究所 | A kind of remote sensing satellite observation mission planing method |
| GB201717125D0 (en) * | 2016-11-28 | 2017-11-29 | National Univ Of Defense Technology | Differential evolution method oriented to agile satellite multi-target task planning |
| CN108038600A (en) * | 2017-12-01 | 2018-05-15 | 中国人民解放军国防科技大学 | Agile earth satellite task planning method |
| CN108052759A (en) * | 2017-12-25 | 2018-05-18 | 航天恒星科技有限公司 | A kind of more star task observation plan method for solving of agility and system based on genetic algorithm |
| CN108388958A (en) * | 2018-01-31 | 2018-08-10 | 中国地质大学（武汉） | A kind of method and device of two-dimensional attitude maneuvering satellite mission planning technical research |
| CN108988933A (en) * | 2018-07-26 | 2018-12-11 | 中国电子科技集团公司第五十四研究所 | A kind of satellite data reception window global optimization distribution method |
-
2019
- 2019-12-24 CN CN201911349972.0A patent/CN111178419B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| GB201717125D0 (en) * | 2016-11-28 | 2017-11-29 | National Univ Of Defense Technology | Differential evolution method oriented to agile satellite multi-target task planning |
| CN107239661A (en) * | 2017-06-05 | 2017-10-10 | 中国电子科技集团公司第五十四研究所 | A kind of remote sensing satellite observation mission planing method |
| CN108038600A (en) * | 2017-12-01 | 2018-05-15 | 中国人民解放军国防科技大学 | Agile earth satellite task planning method |
| CN108052759A (en) * | 2017-12-25 | 2018-05-18 | 航天恒星科技有限公司 | A kind of more star task observation plan method for solving of agility and system based on genetic algorithm |
| CN108388958A (en) * | 2018-01-31 | 2018-08-10 | 中国地质大学（武汉） | A kind of method and device of two-dimensional attitude maneuvering satellite mission planning technical research |
| CN108988933A (en) * | 2018-07-26 | 2018-12-11 | 中国电子科技集团公司第五十四研究所 | A kind of satellite data reception window global optimization distribution method |
Non-Patent Citations (1)
| Title |
|---|
| 郝会成. 敏捷卫星任务规划问题建模及求解方法研究.哈尔滨工业大学博士学位论文 工程科技Ⅱ辑.2015,(第undefined期),133. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111178419A (en) | 2020-05-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111178419B (en) | Agile remote sensing satellite multi-target task planning method based on task clustering | |
| CN110515708B (en) | Satellite on-orbit autonomous task planning method and system | |
| US5537119A (en) | Method and system for tracking multiple regional objects by multi-dimensional relaxation | |
| CN105095643B (en) | The autonomous mission planning method of imaging satellite towards dynamic environment | |
| CN108846504B (en) | Method and system for optimizing multi-point target task in hypersensitive agile satellite region | |
| CN109933842B (en) | Moving target single-star task planning method based on constraint satisfaction genetic algorithm | |
| CN107886201B (en) | Multi-objective optimization method and device for multi-unmanned aerial vehicle task allocation | |
| CN104063748A (en) | Algorithm for imaging satellite-oriented time-dependent scheduling problem | |
| CN102354288A (en) | Imaging quality priority task scheduling method | |
| Zhao et al. | Agile satellite mission planning via task clustering and double-layer tabu algorithm | |
| CN112149911B (en) | Ultra-agile satellite same-orbit multipoint target in-motion imaging task planning method | |
| CN108415045B (en) | Method and device for searching marine moving target by remote sensing satellite | |
| Hong et al. | UAV search-and-rescue planning using an adaptive memetic algorithm | |
| CN110674978A (en) | Task allocation and path planning method for unmanned workshop transportation system | |
| CN114740899B (en) | Gridding airspace distribution and collaborative search planning method | |
| CN113496103B (en) | Agile satellite multi-target collaborative observation method based on combination of multiple principles and chains | |
| CN114399652B (en) | Multi-index system GEO satellite orbit selection method based on radar chart intelligent matching | |
| CN115407352A (en) | Robot positioning and mapping system and method based on particle filtering and mapping optimization | |
| CN114035913A (en) | Satellite multi-target repeated observation task planning method and terminal | |
| CN113063419B (en) | Unmanned aerial vehicle path planning method and system | |
| CN113128749A (en) | Centralized online planning method for satellite observation network | |
| CN113408063B (en) | Multi-satellite multi-reconnaissance target-oriented track maneuvering optimization method | |
| CN110160525A (en) | A large amount of flights change bit path parallelization calculation method under a kind of hazardous weather based on discrete potential energy field | |
| CN114840029A (en) | Jumping grid searching method considering unmanned aerial vehicle maneuverability constraint | |
| He et al. | Cloud avoidance scheduling algorithm for agile optical satellites |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |