[go: up one dir, main page]
More Web Proxy on the site http://driver.im/

CN114580209A - Multi-satellite cooperative task allocation method and system for non-time-sensitive moving target - Google Patents

Multi-satellite cooperative task allocation method and system for non-time-sensitive moving target Download PDF

Info

Publication number
CN114580209A
CN114580209A CN202210407589.1A CN202210407589A CN114580209A CN 114580209 A CN114580209 A CN 114580209A CN 202210407589 A CN202210407589 A CN 202210407589A CN 114580209 A CN114580209 A CN 114580209A
Authority
CN
China
Prior art keywords
satellite
task
representing
distributed
time
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.)
Pending
Application number
CN202210407589.1A
Other languages
Chinese (zh)
Inventor
刘晓路
杨文沅
黄山
左亚辉
淳洁
何磊
陈英武
杜永浩
吕济民
陈盈果
姚锋
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
National University of Defense Technology
Original Assignee
National University of Defense Technology
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by National University of Defense Technology filed Critical National University of Defense Technology
Publication of CN114580209A publication Critical patent/CN114580209A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2111/00—Details relating to CAD techniques
    • G06F2111/04—Constraint-based CAD
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2111/00—Details relating to CAD techniques
    • G06F2111/06—Multi-objective optimisation, e.g. Pareto optimisation using simulated annealing [SA], ant colony algorithms or genetic algorithms [GA]
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D30/00—Reducing energy consumption in communication networks
    • Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Computational Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Pure & Applied Mathematics (AREA)
  • Mathematical Optimization (AREA)
  • Mathematical Analysis (AREA)
  • General Engineering & Computer Science (AREA)
  • Evolutionary Biology (AREA)
  • Computer Hardware Design (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Operations Research (AREA)
  • Probability & Statistics with Applications (AREA)
  • Evolutionary Computation (AREA)
  • Algebra (AREA)
  • Geometry (AREA)
  • Databases & Information Systems (AREA)
  • Software Systems (AREA)
  • Radio Relay Systems (AREA)

Abstract

The invention provides a multi-satellite cooperative task allocation method facing a non-time-sensitive moving target, which is characterized in that a non-time-sensitive moving target task set is injected to an allocation main satellite and then shared to each sub-satellite, each sub-satellite respectively calculates a visible window of each task, the calculation pressure of the allocation main satellite is reduced, after a satellite set which can be allocated by each task is obtained, a rule is evolved based on GEP, and the optimal allocation rule is used for allocating the task to the most appropriate satellite. Experiments prove that the method can carry out task allocation only by selecting the optimal allocation rule, so that the solving speed is higher, and the allocation efficiency is higher.

Description

Multi-satellite cooperative task allocation method and system for non-time-sensitive moving target
Technical Field
The invention belongs to the field of satellite task planning, and particularly relates to a non-time-sensitive moving target-oriented multi-satellite cooperative task allocation method and system.
Background
The Multi-satellite collaborative task allocation problem (MSCTAP) facing non-time-sensitive moving target tracking aims at achieving efficient collaboration of inter-satellite task planning through on-satellite fast task allocation and achieving fast response to dynamic uncertainty of a moving target.
The MSCTAP problem facing non-time-sensitive moving target tracking belongs to a typical constraint optimization problem, and aims to obtain maximization of the benefit of a constellation scheduling task through rapid and reasonable task allocation. The task is generated by an on-satellite autonomous management module, and generally comprises attribute information such as priority, load attribute, imaging duration, satellite window and the like. For the satellite, different state information exists, including current attitude, working state, on-satellite electricity and solid state information. For a task, in a planning period, there may be more than one transit satellite, and how to realize fast high-quality distribution of the task ensures completion of the task with high success rate.
With the enhancement of on-satellite capability (on-satellite image recognition, computation and communication capability), the autonomy of the satellite is realized. In such a case, the response to the moving object can be achieved more quickly. However, whether facing moving targets or conventional observation targets, on-board satellites are confronted with numerous tasks that are dynamically reached and also result from dynamically generated observation targets on-board the satellites. In the face of a plurality of dynamic uncertainties, the high timeliness and the high quality of solving are difficult to simultaneously consider for solving the autonomous mission planning on the satellite by adopting the traditional operation and research optimization method. In the prior art, a heuristic method is adopted, so that a higher profit solution is difficult to obtain, and an accurate or meta-heuristic algorithm is adopted to improve the solving efficiency. The fast and high-quality solution of the on-satellite task planning is the basis for realizing the response to a plurality of on-satellite uncertain factors and is the urgent need of the on-satellite autonomous collaboration at present.
Disclosure of Invention
The invention provides a multi-satellite cooperative task allocation method and system for a non-time-sensitive moving target, aiming at solving the technical problem of how to quickly allocate tasks to satellites with high quality and ensure the completion of the tasks with high success rate for the non-time-sensitive moving target.
In order to solve the technical problems, the technical scheme adopted by the invention is as follows:
a multi-satellite cooperative task allocation method facing non-time-sensitive moving targets comprises the following steps:
step 1: acquiring a non-time-sensitive moving target task set, a satellite set participating in planning and state parameters;
step 2: the non-time-sensitive moving target task set is noted to a main satellite, and meanwhile the non-time-sensitive moving target task set is shared to each sub-satellite in a satellite set, wherein the main satellite is the satellite with the longest transit time, and the sub-satellite is the satellite of other non-main satellites which are about to transit;
and step 3: each subsatellite calculates the visible window of each task according to the self state and the track information to obtain the visible window of each task of each satellite, and feeds back the calculation result of the visible window to the distribution main star;
and 4, step 4: and the distribution main star distributes the multi-star tasks according to the calculation result of the visible window of each satellite to each task fed back by each subsatellite.
Further, the method for distributing the multi-satellite tasks by the main satellite according to the calculation result of the visible window fed back by each sub-satellite in the step 4 is to construct a multi-satellite cooperative task distribution model facing non-time-sensitive moving target tracking, and solve the multi-satellite cooperative task distribution model to obtain a multi-satellite cooperative task distribution scheme.
Further, the multi-satellite cooperative task allocation model is as follows:
Figure BDA0003602628250000031
Figure BDA0003602628250000032
Figure BDA0003602628250000033
Ωj＝{tski|xij＝1} (4)
Figure BDA0003602628250000034
Figure BDA0003602628250000035
wherein n issatRepresenting the number of satellites to be allocated, ntskRepresenting the number of tasks to be distributed, phi (-) representing the scheduling function of the satellite j, returning scheduling benefit, SjRepresents the state attribute, Ω, of satellite j before allocationjSet of tasks, C, representing satellite j allocationjA set of constraints, Ψ, representing the satellite jjRepresents an optimization objective for satellite j;
xijfor decision variables 0-1, 1 denotes the ith task tskiTo satellite j; v. ofijRepresenting a parameter of 0-1, wherein 1 represents that a visible window exists between the task i and the satellite j, and 0 represents that the visible window does not exist between the task i and the satellite j;
Figure BDA0003602628250000036
representing the task imaging load requirement, wherein 0 is optics and 1 is SAR;
Figure BDA0003602628250000037
representing the satellite load type, 0 is optical, and 1 is SAR;
Figure BDA0003602628250000038
the imaging mode of the task i requires that different working modes correspond to different widths;
Figure BDA0003602628250000039
representing the set of operating modes for satellite j.
Further, the method for solving the multi-satellite cooperative task allocation model comprises the following steps:
step 4.1: screening out a satellite set which can be distributed by each task from the calculation result of each satellite to each task visible window through constraint;
step 4.3: and selecting the optimal satellite to be allocated to the task by using the optimal allocation rule in the satellite set which can be allocated to each task in turn.
Further, the optimal allocation rule is obtained by a rule evolution method based on GEP, and specifically includes:
step 4.3.1: randomly generating an initial population, wherein the initial population comprises npopEach chromosome comprises a plurality of coding genes, each coding gene consists of a head part and a tail part, the first coding of the head part must be a function set, other coding bits in the head part can be the function set or a terminal set, and the tail part must be the terminal set;
step 4.3.2: fitness evaluation is carried out on all chromosome individuals in the population;
step 4.3.3: carrying out individual selection, crossing and variation on the population to form a new population;
step 4.3.4: and selecting the individual with the maximum fitness as the optimal distribution rule to output after the iteration times are met.
Further, the function set is +, -, sin, cos, max, min, and the end point set is a feature vector FV representing the multi-star cooperative task allocation problemij，
Figure BDA0003602628250000041
1)TNjRepresenting the proportion of the distribution task on the satellite j to the sum of the distribution tasks of all the satellites;
Figure BDA0003602628250000042
2)REjthe remaining capacity of the satellite j is in proportion;
Figure BDA0003602628250000043
Egymaxwhich represents the maximum amount of power of the satellite,
Figure BDA0003602628250000044
representing the residual capacity of the satellite;
3)APjon a satellite jThe mean value of the priority of the tasks is distributed,
Figure BDA0003602628250000045
normalized, expressed as:
Figure BDA0003602628250000051
primaxrepresents the maximum task priority;
4)SPjrepresents the standard deviation of the priority of the task distributed on the satellite j;
Figure BDA0003602628250000052
normalized, expressed as:
Figure BDA0003602628250000053
5)ARjrepresenting the average of the imaging time length ratio of the priority of the task distributed on the satellite j;
Figure BDA0003602628250000054
after normalization, the method comprises the following steps:
Figure BDA0003602628250000055
6)SRjrepresenting the priority of the task distributed on the satellite j-the imaging duration ratio standard deviation;
Figure BDA0003602628250000056
7)AOjindicating passage of assigned tasks on satellite jMean value of the top time;
Figure BDA0003602628250000057
after normalization, the method comprises the following steps:
Figure BDA0003602628250000058
9)SOjindicating the standard deviation of the over-top time of the task allocated on the satellite j;
Figure BDA0003602628250000061
normalization is carried out by adopting a relative programming cycle Tplan to obtain a normalized product
Figure BDA0003602628250000062
Figure BDA0003602628250000063
9)WPiRepresenting the priority of a task i to be distributed;
normalized, expressed as:
Figure BDA0003602628250000064
10)WRirepresenting the priority-imaging duration of a task i to be distributed;
Figure BDA0003602628250000065
after normalization, the method comprises the following steps:
Figure BDA0003602628250000066
12)WOijto representThe satellite j executes the moment when the task i to be distributed passes the top;
normalizing by a planning period Tplan to obtain normalized
Figure BDA0003602628250000067
Figure BDA0003602628250000068
12)WLijRepresenting the length ratio of the visible window of the task i to be distributed executed by the satellite j;
WLij＝weij-wbij；
after normalization, the method comprises the following steps:
Figure BDA0003602628250000069
13)WCijrepresenting the conflict degree of the tasks i to be distributed and the tasks j to be planned;
WCij＝Cdij；
Figure BDA00036026282500000610
after normalization, the method comprises the following steps:
Figure BDA0003602628250000071
14)WVijrepresenting the length of a non-overlapped time period of the visible window corresponding to the task i to be distributed and other task distribution visible time windows on the satellite j;
Figure BDA0003602628250000072
the function | | · | | represents the sum of the lengths of the window segments in the set;
after normalization, the method comprises the following steps:
Figure BDA0003602628250000073
15)ASijrepresenting the observation slope statistical mean value of the task i to be distributed and the task distributed on the satellite j;
observed slopes sp of task i and task lilCan be defined as:
Figure BDA0003602628250000074
Figure BDA0003602628250000075
after normalization, the method comprises the following steps:
Figure BDA0003602628250000076
17)SSijindicating the standard deviation of observation slopes of the tasks to be distributed i and the tasks to be distributed j;
Figure BDA0003602628250000077
after normalization, the method comprises the following steps:
Figure BDA0003602628250000078
further, the fitness evaluation method for the chromosome individual in step 4.3.2 is as follows:
step 4.3.2.1, decoding the chromosome individual to obtain a distribution rule tree, and converting the distribution rule tree into a regular arithmetic expression;
step 4.3.2.2, for the nsc scenes generated by training, using the rule expressed by the rule arithmetic expression to distribute the tasks to be distributed of each scene;
4.3.2.3, executing single-satellite task scheduling on the tasks distributed to one satellite;
step 4.3.2.4, calculating the total income of all satellites after task scheduling after the regular tasks are distributed according to the regular arithmetic expression in a scene;
and 4.3.2.5, averaging the total benefits after task scheduling in the nsc scenes, and taking the average as an individual fitness evaluation value.
Further, the individual selection in step 4.3.3 uses a tournament ordering method, in the construction of a new parent population, in proportion pro to the population sizeelElite is retained and n is randomly drawn from the populationtourThe individuals with the maximum replication fitness are placed into a new parent;
the operators of the cross mutation are respectively: one-point crossover operator based on the probability of one-point crossover pc1Judging whether to select intersection or not; two-point crossing operator based on the two-point crossing probability pc2Judging whether to select intersection or not; genetic crossover operator based on genetic crossover probability pgeJudging whether to select intersection or not;
the mutation operators are respectively: the single point mutation operator is based on the single point mutation probability pmsJudging whether mutation is needed, segment inversion operator is to select a segment of code segment randomly in chromosome for inversion according to segment inversion probability pmiJudging whether variation is needed; the IS transposition operator randomly selects a segment from the coding segment to insert into a non-first position of another gene of the chromosome according to the transposition probability pisCarrying out the following steps; the RIS transposer randomly selects a fragment starting from the function set from the head of a gene and inserts it into the head starting position of another gene.
The invention also provides a multi-satellite cooperative task allocation system facing the non-time-sensitive moving target, which comprises the following modules:
an information acquisition module: the system comprises a satellite set, a satellite acquisition unit and a data processing unit, wherein the satellite set is used for acquiring a non-time-sensitive moving target task set and participating in planning;
the information uploading module: the system comprises an information acquisition module, a satellite selection module and a satellite selection module, wherein the information acquisition module is used for acquiring a non-time-sensitive moving target task set, and the non-time-sensitive moving target task set is acquired by the information acquisition module and is used for being annotated to a main satellite and sharing the non-time-sensitive moving target task set to each sub-satellite;
the task visible window calculation module: each subsatellite in the satellite set calculates each task visible window in the task set according to the self state and the orbit information to obtain the visible window of each satellite for each task, and feeds back the calculation result of the visible window to the distribution main satellite;
a task allocation module: and the multi-satellite task allocation method is used for allocating the multi-satellite tasks according to the calculation results of the visible windows of each satellite fed back by each subsatellite to each task by the allocation main satellite.
By adopting the technical scheme, the invention has the following beneficial effects:
according to the non-time-sensitive moving target-oriented multi-satellite cooperative task allocation method and system, the non-time-sensitive moving target task set is injected to the allocation main satellite and then shared to the subsategories, the subsategories respectively calculate the visible window of each task, the calculation pressure of the allocation main satellite is reduced, after the satellite set which can be allocated to each task is obtained, the rule is evolved based on GEP, and the optimal allocation rule is used for allocating the tasks to the most appropriate satellite. Experiments prove that the method can carry out task allocation only by selecting the optimal allocation rule, so that the solving speed is higher, and the allocation efficiency is higher.
Drawings
FIG. 1 is a flow chart of the system of the present invention;
fig. 2 is a schematic diagram of a multi-satellite collaboration flow under a centralized-distributed collaboration architecture;
FIG. 3 is a multi-star task allocation problem under serialized decision-making;
FIG. 4 is a MSCTAP problem solving framework for non-time-sensitive moving target tracking;
FIG. 5 is a schematic diagram of a chromosome encoding and decoding process;
FIG. 6 is a schematic diagram of a sequence deconstruction process for task assignment;
FIG. 7 is a flowchart of the evolution of rules based on GEP;
FIG. 8 is a schematic diagram of population individual selection;
FIG. 9 is a diagram of the evolution of GEP at three different scales.
Detailed Description
The technical solutions of the present invention will be described clearly and completely with reference to the accompanying drawings, and it should be understood that the described embodiments are some, but not all embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The Multi-satellite collaborative task allocation problem (MSCTAP) facing non-time-sensitive moving target tracking aims at achieving efficient collaboration of inter-satellite task planning through on-satellite fast task allocation and achieving fast response to dynamic uncertainty of a moving target. The MSCTAP problem facing non-time-sensitive moving target tracking belongs to a typical constraint optimization problem, and aims to obtain maximization of the benefit of a constellation scheduling task through rapid and reasonable task allocation. The enhancement of on-board capabilities (on-board image recognition, computation and communication capabilities) makes satellite autonomy a reality. In such a case, the response to the moving object can be achieved more quickly. However, whether facing moving targets or conventional observation targets, on-board satellites are confronted with numerous tasks that are dynamically reached and also result from dynamically generated observation targets on-board the satellites. In the face of a plurality of dynamic uncertainties, the high timeliness and the high quality of solving are difficult to simultaneously consider for solving the autonomous mission planning on the satellite by adopting the traditional operation and research optimization method. In practice, it is difficult to obtain a higher profit solution by adopting a heuristic method, and solution efficiency needs to be improved by adopting an accurate or meta-heuristic algorithm. The fast and high-quality solution of the on-satellite task planning is the basis for realizing the response to a plurality of on-satellite uncertain factors and is the urgent need of the on-satellite autonomous collaboration at present.
In order to solve the problem of the present invention, some assumptions are made. The method specifically comprises the following steps:
1) the ground satellite observation system has the advantages that the platform carrying visible light loads and SAR loads, which is considered, has agile three-axis attitude maneuvering capability no matter which kind of loads is carried, and can form a section of observation window for a target.
2) The same satellite can only work with one load at the same time, and for the SAR load, the satellite can only be in one imaging mode at each time.
3) The satellite fix constraints are ignored. After the satellite has the image recognition capability, the invalid image can be automatically erased, and the valid image can be compressed. Thus, the retention constraint will no longer be a tight constraint.
4) The electric quantity consumption is generated when the satellite performs imaging and the attitude maneuver, and the electric quantity consumption is respectively in direct proportion to the imaging duration and the attitude maneuver duration.
5) The satellite images the target in a uniform ground speed imaging mode, the observation time of the task is in direct proportion to the length of a generated strip (the side length of a grid) of the task, and the observation time and the length of the generated strip are considered when the task is generated.
6) Neglecting the influence of the imaging observation angle of the satellite on the target on the imaging resolution.
Fig. 1 to fig. 9 show a specific embodiment of a non-time-sensitive moving object oriented multi-satellite cooperative task allocation method according to the present invention, and as shown in fig. 1 and fig. 4, the method includes the following steps:
step 1: acquiring a non-time-sensitive moving target task set, a satellite set participating in planning and state parameters; in this embodiment, the task attributes in the non-time-sensitive moving target task set include attribute information such as a priority, a load attribute, an imaging duration, and a satellite window. For the satellite participating in the planning, the satellite attributes include the current attitude, the working state, the on-satellite electric quantity, the solid memory and other state information. For a task, task allocation means how to realize fast high-quality allocation of the task and ensure completion of the task with high success rate in a planning period, wherein more than one transit satellite may exist in the planning period.
Step 2: and the non-time-sensitive moving target task set is noted to a main satellite, and the non-time-sensitive moving target task set is shared to all sub-satellites, wherein the main satellite is the satellite with the longest transit time, and the sub-satellites are the satellites of other non-main satellites which are about to transit.
In the embodiment, the powerful capability of constellation distributed computation and the capability of on-satellite centralized task allocation for realizing global optimization are utilized. As shown in fig. 2, firstly, the satellite can be annotated with target information of multi-source situation fusion based on the satellite-ground measurement and control network, and the satellite with the longest transit time (corresponding medium orbit satellite) (referred to as the main satellite in this embodiment) serves as a global target maintenance satellite, and meanwhile, the target information is shared but other satellites are not responsible for maintaining and managing the target. And the main star distributes the target information of the tracking observation task to each child star which is about to cross the border.
And step 3: and each subsatellite calculates the visible window of each task according to the self state and the track information to obtain the visible window of each task of each satellite, and feeds back the calculation result of the visible window to the distribution main star.
In this embodiment, the method for calculating the visible window of each task by each sub-satellite according to the self-state and the orbit information is to use STK for calculation, or a method for calculating the visible window is provided in patent document CN105893659A, which is a method for quickly calculating satellite access forecast.
And 4, step 4: and the distribution main star distributes the multi-star tasks according to the calculation result of the visible window of each satellite to each task fed back by each subsatellite. In the embodiment, the main satellite obtains the global information, so that the distribution scheme is better in quality, each sub-satellite obtains the tasks distributed by the main satellite for autonomous scheduling, the observation income sum is maximized, and the result of executing the observation is fed back to the main satellite for distribution. The centralized-distributed cooperative architecture adopted by the invention takes the global information advantage of centralized task allocation and the efficiency of distributed scheduling calculation into account.
In this embodiment, the method for allocating the multi-satellite tasks by the allocating main satellite according to the calculation result of the visible window fed back by each sub-satellite is to construct a multi-satellite cooperative task allocation model facing non-time-sensitive moving target tracking, and solve the multi-satellite cooperative task allocation model to obtain a multi-satellite cooperative task allocation scheme.
The multi-satellite cooperative task allocation model is as follows:
Figure BDA0003602628250000121
Figure BDA0003602628250000122
Figure BDA0003602628250000131
Ωj＝{tski|xij＝1} (4)
Figure BDA0003602628250000132
Figure BDA0003602628250000133
wherein n issatRepresenting the number of satellites to be allocated, ntskRepresenting the number of tasks to be distributed, phi (-) representing the scheduling function of the satellite j, returning scheduling benefit, SjRepresents the state attribute, Ω, of the satellite j before it is assignedjSet of tasks, C, representing satellite j allocationjA set of constraints, Ψ, representing the satellite jjRepresents an optimization objective for satellite j;
xijfor decision variables 0-1, 1 denotes the ith task tskiTo satellite j; v. ofijRepresenting a parameter of 0-1, wherein 1 represents that a visible window exists between the task i and the satellite j, and 0 represents that the visible window does not exist between the task i and the satellite j;
Figure BDA0003602628250000134
representing the task imaging load requirement, wherein 0 is optics and 1 is SAR;
Figure BDA0003602628250000135
representing the satellite load type, 0 is optical, and 1 is SAR;
Figure BDA0003602628250000136
the imaging mode of the task i requires that different working modes correspond to different widths;
Figure BDA0003602628250000137
representing the set of operating modes for satellite j.
In this embodiment, the method for solving the multi-satellite cooperative task allocation model is as follows:
step 4.1: screening satellite sets which can be distributed by each task from the calculation result of each satellite on each task visible window through constraint;
step 4.2: and selecting the optimal satellite from the satellite sets which can be allocated to each task in turn by using the optimal allocation rule to allocate the optimal satellite to the task.
When a non-time-sensitive moving target newly generates a task, the previous task becomes a past formula and fails even if the previous task is not completed, complete re-planning needs to be executed, and the solution efficiency and quality of the algorithm are very important. The multi-satellite cooperative task allocation oriented to non-time-sensitive moving target tracking aims at providing input data for single-satellite complete rescheduling quickly through on-satellite quick allocation. The rapidness and high quality are key points of multi-satellite cooperative task allocation and are also key points for realizing rapid response to non-time-sensitive moving targets. Since the MSCTAP problem facing the non-time-sensitive moving target tracking can be classified as a classical node matching problem, the solving space complexity is exponential. For one with ntskTask and nsatFor the MSCTAP problem facing the non-time-sensitive moving target tracking, constraints such as visible time windows and the like are removed, and n exists in the matching selection of each tasksatThe spatial complexity of the problem is thus
Figure BDA0003602628250000141
And the spatial complexity of the problem is task dependentThe satellite scale growth shows explosive growth, an optimal solution is difficult to obtain in a short time by adopting an accurate solution algorithm, and the black box functions phi (·) cannot be processed by CPLEX classical optimization solution software. For the existing metaheuristic solving method, the timeliness of solving is still not high. The heuristic method has high timeliness of quick solving and generally poor solving quality. Therefore, the idea of solving the multi-satellite cooperative task allocation model in the invention is to model the problem into a sequence decision problem, and gradually construct an optimal allocation scheme of all tasks and satellites by sequentially deciding the optimal allocation satellite of each task. As shown in fig. 3, the assignment of tasks is considered one by one. In the decision of each step, firstly, an allocable satellite set is screened by constraint (as shown by a dotted line in fig. 3), the allocation rule is used for allocating tasks to be allocated in consideration of the current allocation state, a matching relation is established (as shown by a solid line in fig. 3), and the solution of the MSCTAP problem facing to the non-time-sensitive moving target tracking is realized by the multi-step serialized decision, wherein the most critical problem is how to make the allocation rule and describe the current allocation state.
In this embodiment, it is an important link to select which attributes to weight due to the multiple attributes of the task and the satellite, so as to obtain a good allocation rule for allocating the task. Therefore, the distribution rule making method is to train the multi-satellite task distribution rule based on the genetic expression programming evolution method GEP, the distribution rule obtained through training is used for generating an optimal scheme to obtain scores of satellites distributed to a certain specific task, the satellites with the largest income are finally selected based on the scores to be distributed, and the weighting rule is obtained through GEP evolution. As shown in figure 6 of the drawings,
in the evolution method of the GEP-based multi-satellite cooperative task allocation rule in this embodiment, test scenes of multiple task attributes and satellite states are designed, task and satellite parameters in each test scene are extracted as a problem example, modeling is further performed with the problem example, feature extraction is performed according to the following 16 attributes, and then rule evolution is performed to obtain an optimal allocation rule. The training process of the distribution rule is specifically as follows:
step 4.3.1: randomly generating an initial population, wherein the initial population comprises npopEach chromosome comprises a plurality of coding genes, each coding gene consists of a head part and a tail part, the first coding of the head part must be a function set, other coding bits in the head part can be a function set or a terminal set, and the tail part must be a terminal set. In this embodiment, the function set is +, -, sin, cos, max, and min, and the end point set is a feature vector FV representing the multi-star cooperative task allocation problemij，
Figure BDA0003602628250000151
1)TNjRepresenting the proportion of the distribution task on the satellite j to the sum of the distribution tasks of all the satellites;
Figure BDA0003602628250000152
2)REjthe remaining capacity of the satellite j is in proportion;
Figure BDA0003602628250000153
Egymaxwhich represents the maximum amount of power of the satellite,
Figure BDA0003602628250000154
representing the residual capacity of the satellite;
3)APjrepresenting the assigned task priority mean on satellite j,
Figure BDA0003602628250000155
normalized, expressed as:
Figure BDA0003602628250000156
primaxindicating a maximum task priority;
4)SPjrepresents the standard deviation of the priority of the task distributed on the satellite j;
Figure BDA0003602628250000161
normalized, expressed as:
Figure BDA0003602628250000162
5)ARjrepresenting the average of the imaging time length ratio of the priority of the task distributed on the satellite j;
Figure BDA0003602628250000163
after normalization, the method comprises the following steps:
Figure BDA0003602628250000164
6)SRjrepresenting the priority of the assigned task on the satellite j-the imaging duration ratio standard deviation;
Figure BDA0003602628250000165
7)AOjmeans representing the mean of the over-top time of the assigned task on satellite j;
Figure BDA0003602628250000166
after normalization, the method comprises the following steps:
Figure BDA0003602628250000167
8)SOjindicating the standard deviation of the over-the-top time of the task allocated on the satellite j;
Figure BDA0003602628250000168
normalization is carried out by adopting a relative programming period Tplan to obtain a normalized value
Figure BDA0003602628250000169
Figure BDA00036026282500001610
Feature 1) to feature 8) are satellite state related features.
9)WPiRepresenting the priority of a task i to be distributed;
normalized, expressed as:
Figure BDA0003602628250000171
10)WRirepresenting the priority of a task to be distributed i-imaging time length;
Figure BDA0003602628250000172
after normalization, the following steps are performed:
Figure BDA0003602628250000173
feature 9) and feature 10) are the features of the task to be assigned.
11)WOijRepresenting the moment when the satellite j performs the task i to be distributed;
normalizing by a planning period Tplan to obtain normalized
Figure BDA0003602628250000174
Figure BDA0003602628250000175
12)WLijRepresenting the length ratio of the visible window of the task i to be distributed executed by the satellite j;
WLij＝weij-wbij；
after normalization, the following steps are performed:
Figure BDA0003602628250000176
13)WCijrepresenting the conflict degree of the tasks i to be distributed and the tasks j to be planned;
WCij＝Cdij；
Figure BDA0003602628250000177
after normalization, the following steps are performed:
Figure BDA0003602628250000178
14)WVijrepresenting the length of a non-overlapped time period of the visible window corresponding to the task i to be distributed and other task distribution visible time windows on the satellite j;
Figure BDA0003602628250000181
the function | | · | | represents the sum of the lengths of the window segments in the set;
after normalization, the method comprises the following steps:
Figure BDA0003602628250000182
15)ASijstatistical mean value of observation slope representing tasks to be distributed i and tasks distributed on satellite j；
The observed slope spil of task i and task l may be defined as:
Figure BDA0003602628250000183
Figure BDA0003602628250000184
after normalization, the following steps are performed:
Figure BDA0003602628250000185
16)SSijindicating the standard deviation of observation slopes of the tasks to be distributed i and the tasks to be distributed j;
Figure BDA0003602628250000186
after normalization, the following steps are performed:
Figure BDA0003602628250000187
feature 11) to feature 16) are satellite associated features. In the embodiment, the normalization of the 16 attribute features is to improve the applicability and scale generalization of the evolution rule to the problem. The 16 attributes and function sets are used for gene coding, and a schematic diagram of a chromosome coding and decoding process is shown in FIG. 5.
Step 4.3.2: fitness evaluation is performed on all chromosome individuals in the population.
In this embodiment, the method for evaluating fitness of the chromosome individual includes:
and 4.3.2.1, decoding the chromosome individual to obtain a distribution rule tree, and converting the distribution rule tree into a regular arithmetic expression.
Generated for training step 4.3.2.2nscThe scene is distributed to the tasks to be distributed of each scene by using the rules expressed by the rule arithmetic expression;
4.3.2.3, executing single-satellite task scheduling on the tasks distributed to one satellite; common single-star task scheduling methods include heuristic (rules such as priority ordering and earliest executable time ordering), meta-heuristic (genetic and ant colony) scheduling methods and the like.
4.3.2.4, calculating the total income of all satellites after task scheduling is executed after the regular tasks are distributed according to the regular arithmetic expression in a scene;
and 4.3.2.5, averaging the total benefits after task scheduling in the nsc scenes, and taking the average as an individual fitness evaluation value.
In this embodiment, fitness evaluation is a process of evaluating an evolution rule by using a test scenario under a rule corresponding to a given chromosome individual. Fitness evaluation largely determines the evolution direction of the population, and must be consistent with the optimization goal of the problem. And describing a rule arithmetic expression corresponding to each chromosome by adopting a function EvRule (·), wherein the function outputs a score for matching each task i with the satellite j, and the decision making mode by adopting the rule corresponding to the chromosome is as follows:
Figure BDA0003602628250000191
for n generated by trainingscAnd allocating according to the rule expressed by the arithmetic expression. The assignment employs a sequence matching construction process as shown in fig. 7. And after distribution, adopting single-satellite scheduling to obtain scheduling benefit phi kj (·) of each satellite j in the scene k. For the fitness of the rule, the fitness fu of the average income definition rule u of the test scene is adopted in the invention as shown in the following formula.
Figure BDA0003602628250000201
Step 4.3.3: carrying out individual selection, crossing and variation on the population to form a new population;
in this embodiment, the tournament ordering method is used for individual selection, and when a new parent population is constructed, pro is a proportion of the population sizeelElite is retained and n is randomly drawn from the populationtourAnd (4) putting the individual with the maximum replication fitness into a new parent. The population individuals are selected as new parents of the next generation, and elite individuals need to be reserved, and certain inferior individuals need to be reserved to avoid falling into local optimality. As shown in fig. 8.
The crossover operator is designed to maintain population diversity and enables the search for rules by exchanging between chromosome coding segments. The crossover operator in this embodiment employs:
one-point crossover operator based on the probability of one-point crossover pc1Judging whether to select crossover, namely a classical 1-opt operator;
two-point crossing operator based on the two-point crossing probability pc2Judging whether to select crossover, namely a classical 2-opt operator;
genetic crossover operator based on the probability of genetic crossover pgeIt is determined whether a crossover is required.
The mutation operator is also used for realizing regular search and maintaining population diversity, and compared with a crossover operator which acts on two chromosomes, the mutation operator acts on a single chromosome. The mutation operator in this embodiment employs:
single point mutation operator: according to the probability of single point variation pmsJudging whether variation is needed;
segment inversion operator, which randomly selects a segment of code segment in chromosome for inversion according to segment inversion probability pmiJudging whether variation is needed;
the IS transposition operator randomly selects a segment from the coding segment to insert into a non-first position of another gene of the chromosome according to the transposition probability pisCarrying out the following steps;
the RIS transposer randomly selects a fragment starting from the function set from the head of a gene and inserts it into the head starting position of another gene.
Step 4.3.4: and selecting the individual with the maximum fitness as the optimal distribution rule to output after the iteration times are met.
And (3) experimental verification:
aiming at the MSCTAP problem oriented to non-time-sensitive moving target tracking, the invention aims to design a fast, efficient and excellent solving algorithm so as to achieve the fast response to the dynamic change of a moving target and realize the fast distribution of on-satellite tasks. The test platform is desktop (CPU: Inter (R), core (TM) i5-4460, main frequency 3.2GHz and RAM: 12 GB).
And the evolution process analysis mainly comprises the step of checking the convergence condition and the evolution result of the GEP evolution process. In the evolution process, the experiment is regularly evolved for three task scales, namely sc100-s4-t160, sc100-s7-t280 and sc100-s10-t 400. These three types of scales respectively represent the number of scenes nscThe number of the satellites is 100, the number of the satellites is 4, and the number of tasks to be distributed is 160; the number of satellites is 7, and the number of tasks to be distributed is 280; and the scene scale with the satellite number of 10 and the number of tasks to be distributed of 400 respectively represents the regular evolution of small, medium and large-scale scenes. During the evolution process, the parameter settings of the algorithm are shown in table 1.
Through the fitness evaluation scenes set in three scales, the maximum fitness, the minimum fitness and the average fitness of each generation of individuals in the evolution process are counted through experiments, the final statistical result of the evolution process is obtained and is shown in fig. 9, and as can be seen from fig. 9, the fitness value of the scenes in the three scales rises very fast in the previous 50 generations, which shows that the algorithm has an obvious effect on improving the regular evolution in the early stage. For small-scale scenes, the evolution algebra reaches 50 and then gradually slows down, converges roughly around 200 generations, and then presents a steady trend of stopping growth. The medium and large scales, while tending to plateau after 100 generations, lifted slightly after 350 and 300 generations respectively, then stopped growing and tended to plateau. In the whole evolution process, the fitness average value of the population keeps the same trend with the maximum value, but the minimum value is basically in a fluctuation stage, so that the population diversity is favorably maintained, and the population is prevented from falling into local optimum. Through experimental analysis, the convergence and the trend consistency of the evolution rule are guaranteed in the GEP algorithm. The evolution effect of the GEP on the MSCTAP problem solving rule facing the non-time-sensitive moving target tracking is obvious.
TABLE 1 GEP evolution Process parameters
Parameter name Parameter value Parameter name Parameter value
Population size npop 50 Evolution algebra niter 600
Fitness evaluation scene number n sc 100 Probability of single point crossing pcl 0.7
Probability of double point crossing pc2 0.7 Probability of gene crossing pge 0.7
Probability of single point crossing pms 0.05 Probability of segment inversion pmi 0.1
IS transposition probability pis 0.1 RIS transposition probability pris 0.1
Elite individual retention ratio proel 0.1 Sorting and extracting the number n of individualstour 3
The satellite is in transit for a period of time during which the task has a window to be likely assigned to the satellite. According to the index of the task allocation number allocated to the satellite on each bundle of task windows, except that the number of tasks allocated to the satellite 3 in 33 and the number of tasks allocated to the satellite 9 in 34 in a large-scale scene are relatively small, the number of tasks allocated to the satellite in other large-scale scenes is basically maintained at about 40, and not only is the distribution uniformity, but also the conflict between the tasks is reduced for each satellite in the single-satellite task scheduling. If the distribution is lack of uniformity, higher conflict will occur on a certain satellite, the success rate of task scheduling is reduced, and the reduction of the overall benefit of multi-satellite task planning is brought. In addition, from the uniformity of the distribution of task allocation windows, the distribution of the task allocation windows on each satellite is relatively uniform, and the phenomenon of excessive accumulation at a certain section of the satellite window does not occur, which shows that the overlapping degree of the windows between tasks and the tasks allocated by the allocation satellites can be reduced as much as possible when the allocation is performed by the evolution rule, and the conflict between the tasks is reduced. This trend in task allocation clearly increases the likelihood of task scheduling success. In addition, the distribution time of the experiment under three scale example scenes is counted, and is respectively 2.324s, 6.744s and 13.199s, and from the actual use condition, the timeliness of the algorithm distribution meets the satellite autonomous requirement and can respond to the uncertainty caused by the motion change of a non-time-sensitive moving target. The above analysis can find that a Heuristic multi-satellite cooperative task allocation Method CHMGEP (structured statistical Method for multi-satellite cooperative task allocation on GEP evolution, CHMGEP) is feasible in solving the multi-satellite task allocation problem.
In addition, 50 calculation examples are generated for scenes with the scales of s4-t160, s7-t280 and s10-t400 respectively and are tested, and finally, scheduling benefit (SP) and Scheduling Time (ST) in each scene are counted, it should be noted that the scheduling time includes the time of multi-satellite allocation time and single-satellite scheduling, and the single-satellite scheduling time is serial (in a centralized-distributed collaborative architecture of research design, the single-satellite scheduling may be parallel, and therefore, the calculation time may be greatly reduced in practical application). Through experiments, the CHMGEP algorithm realizes comprehensive surpassing on SP indexes of other three heuristic algorithms under three scene algorithms of small, medium and large scales by counting the test results of three 50-scale algorithms. The three heuristic algorithms respectively have advantages and disadvantages in solving the problem quality, and the expression of the LVTD algorithm is long. It can be seen that the CHMGEP algorithm is advantageous over a single heuristic algorithm.
Finally, it should be noted that: the above embodiments are only used to illustrate the technical solution of the present invention, and not to limit the same; while the invention has been described in detail and with reference to the foregoing embodiments, it will be understood by those skilled in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some or all of the technical features may be equivalently replaced; and the modifications or the substitutions do not make the essence of the corresponding technical solutions depart from the scope of the technical solutions of the embodiments of the present invention.

Claims (9)

1. A multi-satellite cooperative task allocation method facing a non-time-sensitive moving target is characterized by comprising the following steps:
step 1: acquiring a non-time-sensitive moving target task set, a satellite set participating in planning and state parameters;
step 2: the non-time-sensitive moving target task set is noted to a main satellite, and meanwhile the non-time-sensitive moving target task set is shared to each sub-satellite in a satellite set, wherein the main satellite is the satellite with the longest transit time, and the sub-satellite is the satellite of other non-main satellites which are about to transit;
and step 3: each subsatellite calculates the visible window of each task according to the self state and the track information to obtain the visible window of each task of each satellite, and feeds back the calculation result of the visible window to the distribution main star;
and 4, step 4: and the distribution main star distributes the multi-star tasks according to the calculation result of the visible window of each satellite to each task fed back by each subsatellite.
2. The distribution method according to claim 1, wherein the method for distributing the multi-satellite tasks by the main satellite according to the calculation result of the visible window fed back by each sub-satellite in the step 4 is to construct a multi-satellite cooperative task distribution model facing non-time-sensitive moving target tracking, and solve the multi-satellite cooperative task distribution model to obtain a multi-satellite cooperative task distribution scheme.
3. The distribution method according to claim 2, wherein the multi-satellite cooperative task distribution model is:
Figure FDA0003602628240000011
Figure FDA0003602628240000012
Figure FDA0003602628240000021
Ωj＝{tski|xij＝1} (4)
Figure FDA0003602628240000022
Figure FDA0003602628240000023
wherein n issatRepresenting the number of satellites to be allocated, ntskRepresenting the number of tasks to be distributed, phi (-) representing the scheduling function of the satellite j, returning scheduling benefit, SjRepresents the state attribute, Ω, of the satellite j before it is assignedjSet of tasks, C, representing satellite j allocationjA set of constraints, Ψ, representing the satellite jjRepresents an optimization objective for satellite j;
xijfor decision variables 0-1, 1 denotes the ith task tskiTo satellite j; v. ofijRepresenting a parameter of 0-1, wherein 1 represents that a visible window exists between the task i and the satellite j, and 0 represents that the visible window does not exist between the task i and the satellite j;
Figure FDA0003602628240000024
representing the task imaging load requirement, wherein 0 is optics and 1 is SAR;
Figure FDA0003602628240000025
representing the satellite load type, 0 is optical, and 1 is SAR;
Figure FDA0003602628240000026
the imaging mode of the task i requires that different working modes correspond to different widths;
Figure FDA0003602628240000027
representing the set of operating modes for satellite j.
4. The distribution method according to claim 3, wherein the method for solving the multi-satellite cooperative task distribution model is as follows:
step 4.1: screening satellite sets which can be distributed by each task from the calculation result of each satellite on each task visible window through constraint;
step 4.3: and selecting the optimal satellite to be distributed to the task by using a distribution rule in the satellite set which can be distributed to each task in turn.
5. The distribution method according to claim 4, wherein the distribution rule is obtained by a GEP-based multi-star distribution rule evolution method, and specifically comprises:
step 4.3.1: randomly generating an initial population, wherein the initial population comprises npopEach chromosome comprises a plurality of coding genes, each coding gene consists of a head part and a tail part, the first coding of the head part must be a function set, other coding bits in the head part can be the function set or a terminal set, and the tail part must be the terminal set;
step 4.3.2: fitness evaluation is carried out on all chromosome individuals in the population;
step 4.3.3: carrying out individual selection, crossing and variation on the population to form a new population;
step 4.3.4: and selecting the individual with the maximum fitness as the optimal distribution rule to output after the iteration times are met.
6. The assignment method according to claim 5, wherein the function set is +, -, sin, cos, max, min, and the end point set is a feature vector FV characterizing the multi-star collaborative task assignment problemij，
Figure FDA0003602628240000031
1)TNjIndicating assigned mission accommodation on satellite jThere is a proportion of the total of satellite allocation tasks;
Figure FDA0003602628240000032
2)REjthe remaining capacity of the satellite j is in proportion;
Figure FDA0003602628240000033
Egymaxwhich represents the maximum amount of power of the satellite,
Figure FDA0003602628240000034
representing the residual capacity of the satellite;
3)APjrepresenting the assigned task priority mean on satellite j,
Figure FDA0003602628240000035
normalized, expressed as:
Figure FDA0003602628240000036
primaxindicating a maximum task priority;
4)SPjrepresents the standard deviation of the priority of the task distributed on the satellite j;
Figure FDA0003602628240000041
normalized, expressed as:
Figure FDA0003602628240000042
5)ARjrepresenting the average of the imaging time length ratio of the priority of the task distributed on the satellite j;
Figure FDA0003602628240000043
after normalization, the method comprises the following steps:
Figure FDA0003602628240000044
6)SRjrepresenting the priority of the assigned task on the satellite j-the imaging duration ratio standard deviation;
Figure FDA0003602628240000045
7)AOjmeans representing the mean of the over-top time of the assigned task on satellite j;
Figure FDA0003602628240000046
after normalization, the method comprises the following steps:
Figure FDA0003602628240000047
8)SOjindicating the standard deviation of the over-top time of the task allocated on the satellite j;
Figure FDA0003602628240000048
using a relative programming period TplanNormalization is carried out to obtain normalized
Figure FDA0003602628240000049
Figure FDA0003602628240000051
9)WPiRepresenting the priority of a task i to be distributed;
normalized, expressed as:
Figure FDA0003602628240000052
10)WRirepresenting the priority of a task to be distributed i-imaging time length;
Figure FDA0003602628240000053
after normalization, the method comprises the following steps:
Figure FDA0003602628240000054
11)WOijrepresenting the moment when the satellite j performs the task i to be distributed;
by planning the period TplanNormalization is carried out to obtain normalized
Figure FDA0003602628240000055
Figure FDA0003602628240000056
12)WLijRepresenting the length ratio of the visible window of the task i to be distributed executed by the satellite j;
WLij＝weij-wbij；
after normalization, the method comprises the following steps:
Figure FDA0003602628240000057
13)WCijrepresenting the conflict degree of the tasks i to be distributed and the tasks j to be planned;
WCij＝Cdij；
Figure FDA0003602628240000058
after normalization, the method comprises the following steps:
Figure FDA0003602628240000059
14)WVijrepresenting the length of a non-overlapped time period of the visible window corresponding to the task i to be distributed and other task distribution visible time windows on the satellite j;
Figure FDA0003602628240000061
the function | | · | | represents the sum of the lengths of the window segments in the set;
after normalization, the method comprises the following steps:
Figure FDA0003602628240000062
15)ASijrepresenting the observation slope statistical mean value of the task i to be distributed and the task distributed on the satellite j;
observed slopes sp of task i and task lilCan be defined as:
Figure FDA0003602628240000063
Figure FDA0003602628240000064
after normalization, the method comprises the following steps:
Figure FDA0003602628240000065
16)SSijindicating the standard deviation of observation slopes of the tasks to be distributed i and the tasks to be distributed j;
Figure FDA0003602628240000066
after normalization, the method comprises the following steps:
Figure FDA0003602628240000067
7. the assignment method according to claim 6, wherein the fitness evaluation of the individual chromosome in step 4.3.2 is performed by:
step 4.3.2.1, decoding the chromosome individual to obtain a distribution rule tree, and converting the distribution rule tree into a regular arithmetic expression;
step 4.3.2.2, for the nsc scenes generated by training, using the rule expressed by the rule arithmetic expression to distribute the tasks to be distributed of each scene;
4.3.2.3, executing single-satellite task scheduling on the tasks distributed to one satellite;
step 4.3.2.4, calculating the total income of all satellites after task scheduling after the regular tasks are distributed according to the regular arithmetic expression in a scene;
step 4.3.2.5, until n is obtainedscAnd (4) calculating the total income after task scheduling in each scene, and averaging the total income to serve as an individual fitness evaluation value.
8. The distribution method according to claim 6, characterized in that the individual selection in step 4.3.3 uses a tournament ordering method, in the construction of a new parent population, in a proportion pro of the population sizeelElite is retained, followed by the populationMachine extraction ntourThe individual with the maximum fitness is copied and put into a new parent;
the operators of the cross mutation are respectively: one-point crossover operator based on the probability of one-point crossover pc1Judging whether to select intersection or not; two-point crossing operator based on the two-point crossing probability pc2Judging whether to select intersection or not; genetic crossover operator based on the probability of genetic crossover pgeJudging whether to select intersection or not;
the mutation operators are respectively: the single point mutation operator is based on the single point mutation probability pmsJudging whether mutation is needed, segment inversion operator is to select a segment of code segment randomly in chromosome for inversion according to segment inversion probability pmiJudging whether variation is needed; the IS transposition operator randomly selects a segment from the coding segment to insert into a non-first position of another gene of the chromosome according to the transposition probability pisCarrying out the following steps; the RIS transposer randomly selects a fragment starting from the function set from the head of a gene and inserts it into the head starting position of another gene.
9. A multi-satellite cooperative task distribution system facing a non-time-sensitive moving target is characterized by comprising the following modules:
an information acquisition module: the system comprises a satellite set, a satellite acquisition unit and a data processing unit, wherein the satellite set is used for acquiring a non-time-sensitive moving target task set and participating in planning;
the information uploading module: the system comprises an information acquisition module, a satellite selection module and a satellite selection module, wherein the information acquisition module is used for acquiring a non-time-sensitive moving target task set, and the non-time-sensitive moving target task set is acquired by the information acquisition module and is used for being annotated to a main satellite and sharing the non-time-sensitive moving target task set to each sub-satellite;
the task visible window calculation module: each subsatellite in the satellite set calculates each task visible window in the task set according to the self state and the orbit information to obtain the visible window of each satellite for each task, and feeds back the calculation result of the visible window to the distribution main satellite;
a task allocation module: and the multi-satellite task allocation method is used for allocating the multi-satellite tasks according to the calculation results of the visible windows of each satellite fed back by each subsatellite to each task by the allocation main satellite.
CN202210407589.1A 2022-01-13 2022-04-19 Multi-satellite cooperative task allocation method and system for non-time-sensitive moving target Pending CN114580209A (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202210035158 2022-01-13
CN2022100351587 2022-01-13

Publications (1)

Publication Number Publication Date
CN114580209A true CN114580209A (en) 2022-06-03

Family

ID=81779429

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210407589.1A Pending CN114580209A (en) 2022-01-13 2022-04-19 Multi-satellite cooperative task allocation method and system for non-time-sensitive moving target

Country Status (1)

Country Link
CN (1) CN114580209A (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117271146A (en) * 2023-11-23 2023-12-22 中国人民解放军战略支援部队航天工程大学 Multi-star imaging task planning method based on knowledge transfer under complex heterogeneous scene
CN117872770A (en) * 2024-01-24 2024-04-12 中国人民解放军军事科学院国防科技创新研究院 Satellite cluster task decision and distribution method

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117271146A (en) * 2023-11-23 2023-12-22 中国人民解放军战略支援部队航天工程大学 Multi-star imaging task planning method based on knowledge transfer under complex heterogeneous scene
CN117271146B (en) * 2023-11-23 2024-02-23 中国人民解放军战略支援部队航天工程大学 Multi-star imaging task planning method based on knowledge transfer under complex heterogeneous scene
CN117872770A (en) * 2024-01-24 2024-04-12 中国人民解放军军事科学院国防科技创新研究院 Satellite cluster task decision and distribution method

Similar Documents

Publication Publication Date Title
CN112766813B (en) Space-day collaborative observation complex task scheduling method
CN109165808B (en) Power communication network on-site operation and maintenance work order distribution method
CN108388958B (en) Method and device for researching two-dimensional attitude maneuvering satellite mission planning technology
CN113885555B (en) Multi-machine task allocation method and system for transmission line dense channel inspection
CN109960544B (en) Task parallel scheduling method based on data driving type agile satellite
CN114580209A (en) Multi-satellite cooperative task allocation method and system for non-time-sensitive moving target
Khmeleva et al. Fuzzy-logic controlled genetic algorithm for the rail-freight crew-scheduling problem
CN109767128B (en) Imaging satellite autonomous task planning method based on machine learning
CN109409773B (en) Dynamic planning method for earth observation resources based on contract network mechanism
Ribas et al. A micro-genetic algorithm for multi-objective scheduling of a real world pipeline network
Abedinia et al. A new reconfigured electricity market bidding strategy in view of players' concerns
Seydanlou et al. A customized multi-neighborhood search algorithm using the tabu list for a sustainable closed-loop supply chain network under uncertainty
Turan et al. A multi-objective simulation–optimization for a joint problem of strategic facility location, workforce planning, and capacity allocation: A case study in the Royal Australian Navy
Ebrahimnejad A method for solving linear programming with interval-valued trapezoidal fuzzy variables
CN115421885B (en) Distributed multi-target cloud task scheduling method and device and cloud service system
CN104537446A (en) Bilevel vehicle routing optimization method with fuzzy random time window
CN115271130A (en) Dynamic scheduling method and system for maintenance order of ship main power equipment
CN115511336A (en) Shared single-vehicle parking spot addressing method and system based on multi-point joint distribution
CN104392317A (en) Project scheduling method based on genetic culture gene algorithm
Turan et al. Integrating decision maker preferences to a risk-averse multi-objective simulation-based optimization for a military workforce planning, asset management and fleet management problem
CN110807543A (en) Investment portfolio optimization method and device based on group decision intelligent search
Zhang et al. A new evolutionary algorithm based on MOEA/D for portfolio optimization
CN112966873B (en) Industrial park comprehensive energy system optimization operation method, device and storage medium
CN110689320A (en) Large-scale multi-target project scheduling method based on co-evolution algorithm
Zhang et al. A novel collaborative agent reinforcement learning framework based on an attention mechanism and disjunctive graph embedding for flexible job shop scheduling problem

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