CN108377163A - Multi-satellite online collaboration method based on asynchronous communication - Google Patents

Multi-satellite online collaboration method based on asynchronous communication Download PDF

Info

Publication number
CN108377163A
CN108377163A CN201711276877.3A CN201711276877A CN108377163A CN 108377163 A CN108377163 A CN 108377163A CN 201711276877 A CN201711276877 A CN 201711276877A CN 108377163 A CN108377163 A CN 108377163A
Authority
CN
China
Prior art keywords
task
asynchronous
satellite
communication
ascl
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.)
Granted
Application number
CN201711276877.3A
Other languages
Chinese (zh)
Other versions
CN108377163B (en
Inventor
陈英武
李国梁
陈宇宁
吕济民
陈盈果
陈成
王涛
刘晓路
邢立宁
姚锋
贺仁杰
张忠山
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
Priority to CN201711276877.3A priority Critical patent/CN108377163B/en
Publication of CN108377163A publication Critical patent/CN108377163A/en
Application granted granted Critical
Publication of CN108377163B publication Critical patent/CN108377163B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04B—TRANSMISSION
    • H04B7/00—Radio transmission systems, i.e. using radiation field
    • H04B7/14—Relay systems
    • H04B7/15—Active relay systems
    • H04B7/185—Space-based or airborne stations; Stations for satellite systems
    • H04B7/1851—Systems using a satellite or space-based relay
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04B—TRANSMISSION
    • H04B7/00—Radio transmission systems, i.e. using radiation field
    • H04B7/14—Relay systems
    • H04B7/15—Active relay systems
    • H04B7/185—Space-based or airborne stations; Stations for satellite systems
    • H04B7/1851—Systems using a satellite or space-based relay
    • H04B7/18519—Operations control, administration or maintenance
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04B—TRANSMISSION
    • H04B7/00—Radio transmission systems, i.e. using radiation field
    • H04B7/14—Relay systems
    • H04B7/15—Active relay systems
    • H04B7/185—Space-based or airborne stations; Stations for satellite systems
    • H04B7/18521—Systems of inter linked satellites, i.e. inter satellite service

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Astronomy & Astrophysics (AREA)
  • Aviation & Aerospace Engineering (AREA)
  • General Physics & Mathematics (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Radio Relay Systems (AREA)

Abstract

The invention discloses a multi-satellite online cooperation method based on asynchronous communication, which adopts an improved consistency bundle algorithm based on synchronous communication, aims at the complex combination optimization problem of multi-satellite online cooperation task scheduling under communication constraint, and is oriented to a distributed architecture.

Description

A kind of online Synergistic method of more stars based on asynchronous communication
Technical field
The present invention relates to satellite technology fields, more particularly to a kind of online Synergistic method of more stars based on asynchronous communication.
Background technology
With the promotion of load detectivity and image-capable on star, earth observation satellite can find have on star The observed object of value and to generate the request of further observation mission (be typically specifically to have answering compared with high yield or priority level Anxious observation mission), the observation mission request (emergent observation mission request) that other satellite communications pass over also is can receive, and The request of these observation missions is that dynamic random reaches and timeliness requires height, needs to solve online, (especially same in more stars The more stars of structure) in the case of, carry out cooperative scheduling to need the online collaborative dispatch mechanism of more stars.
Invention content
The purpose of the present invention is to provide a kind of online Synergistic methods of more stars being based on asynchronous communication to realize isomorphism Observation mission of meeting an urgent need especially preferably is coped in the online collaboration of more stars.
To achieve the above object, the present invention provides a kind of online Synergistic method of more stars based on asynchronous communication, more stars Using decentralized architecture, every earth observation satellite LEO is autonomous satellite, and the intelligence level having the same in framework, no There are grade difference, the autonomous satellite is communicated using geostationary orbit telecommunication satellite as relay node, and ground Ball stationary orbit telecommunication satellite plays the effect of synergist, is responsible for predicting asynchronous communication loop, for emergent urgent sight Survey task, every autonomous satellite all pass through the cyclic process of " contexture by self dispatches --- communication interaction --- Coordination Decision ", In, every autonomous satellite is not only updated the task scheduling of itself, also to the task meter of other autonomous satellites in framework It draws and is safeguarded, the online Synergistic method of more stars based on asynchronous communication uses the asynchronous consistency beam based on asynchronous communication Algorithm.
Preferably, the asynchronous consistency beam algorithm based on asynchronous communication includes：Asynchronous communication loop forecast period, Beam builds the stage and consistency builds the stage, in the asynchronous communication loop forecast period, when earth observation satellite is to new one It criticizes after emergent observation mission is scheduled and its information vector is shared with geostationary orbit telecommunication satellite immediately, describedly Ball stationary orbit telecommunication satellite predicts asynchronous communication loop ACL,
For each earth observation satellite, in the pot life section of ACL, when the task beam of promoter or respondent When changing, then asynchronous feedback of sharing is executed on ACL and is interacted, and is equal downlink is shared with the timestamp of feeding back uplink, tDS≈tUF, every time it is asynchronous share feedback interaction in, carry out beam structure the stage and consistency build the stage, wherein asynchronous point Enjoy feedback interaction be on available asynchronous communication loop trigger generate, it is asynchronous share feedback interaction in, altogether there are four when Between stab, four timestamps include share uplink, share downlink, feeding back uplink and feeding back downlink,
Wherein, asynchronous communication loop ACL is made of two or three independent asynchronous sub- communications loop ASCL, ASCL's Description form is multi-component system ＜ ID, I, R, ES, LF a ＞ASCL,
- ID is an identifier；
- I is communication initiator number, i.e., first is shared its information；
- R is the communication response person number for receiving simultaneously sharing information；
- ES is the earliest start time of asynchronous sub- communications loop；
- LF is the end time the latest of asynchronous sub- communications loop,
I or R is relay node, and when a LEO is determined as I, R is then set as relay node, otherwise, when I is relay node When, then R is set as LEO,
In the case where asynchronous sub- communications loop ASCL has the time interval of overlapping, asynchronous communication loop ACL is by two Independent asynchronous sub- communications loop ASCL is constituted, and does not have the case where time interval of overlapping in asynchronous sub- communications loop ASCL Under, asynchronous communication loop ACL is made of three independent asynchronous sub- communications loop ASCL.
Preferably, the stage is built in beam, task is constantly added to by every earth observation satellite in a manner of succession greediness In its task beam, until cannot add any task again, the task in task beam is ranked sequentially according to what is added successively , and the task in task scheduling is ranked up according to its corresponding practical time started, and the stage is built in beam, each Earth observation satellite all builds respective task beam, and is updated with the progress of scheduling process, appoints for being not currently in The each available task being engaged in beam, earth observation satellite compare its financial value and the task score in current acceptance of the bid vector Compared with if the financial value bigger, new task acceptance of the bid score being left, for unscheduled set of tasks, earth observation Satellite therefrom chooses the task with top score, and is added in task beam, and the recursive procedure of beam structure is held always Continue task beam and have reached its capacity limit, or there is no tasks can be added in task beam, is i.e. the earth observation satellite The execution of remaining task is no longer able to surpass other earth observation satellites.
Preferably, in two asynchronous sub- communications loop ＜ ID1,I1,R1,ES1,LF1＞ASCLWith ＜ ID2,I2,R2,ES2,LF2 ＞ASCLIn the case of forming the asynchronous communication loop, in the two asynchronous sub- communications loops, there are following five kinds of correlations Relationship：
I and R is LEO satellite, and ACL promoters I is the promoter I of previous ASCL1, while ACL respondent R is latter The respondent R of ASCL2；
I=I1
R=R2
The respondent R of previous ASCL1With the promoter I of latter ASCL2It is communication relay node；
When communication relay node is interconnected, then the corresponding relay nodes of two ASCL can differ,
Otherwise, corresponding relay node is same；
ESUSAnd LFUSIndicate that ACL's shares uplink communication time window；
ESUS=ES1
LFUS=min { LF1,LF2}
ESUS≤LFUS
ESDS-UFAnd LFDS-UFIndicate that ACL's shares downlink-feeding back uplink call duration time window；
ESDS-UF=max { ES1,ES2}
LFDS-UF=LF2
ESDS-UF≤LFDS-UF
ESDFAnd LFDFIndicate the feeding back downlink call duration time window of ACL；
ESDF=max { ES1,ES2}
LFDF=LF1
ESDF≤LFDF。
Preferably, in three asynchronous sub- communications loop ＜ ID1,I1,R1,ES1,LF1＞ASCL, ＜ ID2,I2,R2,ES2,LF2 ＞ASCLWith ＜ ID3,I3,R3,ES3,LF3＞ASCLIn the case of forming the asynchronous communication loop, in these three asynchronous sub- communications In loop, there are following correlativities：
I and R is LEO satellite, and the promoter I of ACL is the promoter I of first ASCL1, meanwhile, respondent R is most The respondent R of the latter ASCL3,
I=I1
R=R3
The promoter I of first ASCL1With the respondent R of the last one ASCL3It is identical, to constitute a full communication ring Road,
I1=R3
The respondent R of first ASCL1With the promoter I of latter two ASCL2And I3It is communication relay node, works as relaying When being interconnected between node, then corresponding relay node can be different, are the same relay nodes otherwise,
ESUSAnd LFUSIndicate that ACL's shares uplink communication time window,
ESUS=ES1
LFUS=min { LF1,LF2}
ESUS≤LFUS
ESDS-UFAnd LFDS-UFIndicate that ACL's shares downlink-feeding back uplink call duration time window,
ESDS-UF=max { ES1,ES2}
LFDS-UF=LF2
ESDS-UF≤LFDS-UF
ESDFAnd LFDFIndicate the feeding back downlink call duration time window of ACL,
ESDF=max { ES2,ES3}
LFDF=LF3
ESDF≤LFDF。
Preferably, respondent is using one kind in following five kinds action：
1) update ＆ shares again：The information that respondent sends according to sender updates the person's of winning list, acceptance of the bid bid row Table and time arrow of winning, then share current update again；
2) give up ＆ again to share：Respondent does not change its information state, then to the person's of winning information for itself being assert into Row is shared again；
3) give up ＆ no longer to share：Respondent does not change the information state of itself, does not share again to it yet, applied to being connect The message received the situation consistent with existing information；
4) resetting ＆ shares again：Respondent is zeroed out the person of winning and value of winning, the message that then original is received into Row is shared again, then eliminates confusion；
5) renewal time ＆ shares again：Respondent is the person of winning, and when it receives a message, generated time stamp is won to it more It is newly current time, to confirm that it oneself is the person of winning oneself to still believe that.
Preferably, when being responded to the message sent from the earth observation satellite as sender, appoint for emergent Be engaged in oj, operates decision rule such as following table used by square earth observation satellite in response,
Local decision rule under asynchronous communication when it receives a message
。
Description of the drawings
Fig. 1 is the more star co-architecture block diagrams of distributing.
Fig. 2 is the information flow chart in the more star co-architectures of distributing.
Fig. 3 is the general conception figure that online cooperative scheduling solves on entire scheduling interval.
Fig. 4 is the communication scheme between relay node and LEO satellite.
Fig. 5 a and Fig. 5 b are asynchronous communication loop schematic diagram.
Fig. 6 shows available communication between space and ground segment.
Fig. 7 a- Fig. 7 c are for the m-CBBA arithmetic results not interconnected between GEO with interconnection situation, and wherein Fig. 7 a are shown Total revenue, Fig. 7 b show that the contingency tasks percentage of successful dispatch, Fig. 7 c show total number of communications.
Fig. 8 a- Fig. 8 c be each algorithm task spatial distribution U (- 45,45) and under the different rolling scheduling periods as a result, Wherein Fig. 8 a show that total revenue, Fig. 8 b show that the percentage of contingency tasks successful dispatch, Fig. 8 c show total number of communications.
Specific implementation mode
The embodiment of the present invention is described in detail below in conjunction with the accompanying drawings.Earth observation satellite LEO in the present invention is Autonomous agent (autonomous satellite) has following characteristics：Autonomous perception, contexture by self scheduling are executed with autonomous.The body of independence Now aspect includes determining accurate posture, track determination, attitude sensor/driver and load calibration, gesture stability, track machine Dynamic, Stateful Inspection and forecast analysis, diagnosis, are isolated and reparation, forecast type modeling (are supervised in planned dispatching, state fault detect Determined in holding etc. depending on analysis and track, need the support of forecast type model, such as ephemeris, sun intensity information), task meter It draws and scheduling, load control and configuration, data are stored and communicated and image real time transfer etc..For example, sense can be passed through Know to be autonomously generated emergency scheduling task.
The method of the present invention is particularly suitable for the earth observation satellite of decentralized architecture (also referred to as distributing co-architecture) System.In distributing co-architecture, each autonomous satellite intelligent level having the same is not present grade difference, entire group It is flat and fully distributed to knit.Meanwhile there is communication interaction link between all autonomous satellites.Therefore, which has There are very strong adaptability and reliability, wherein each autonomous satellite can carry out intelligent decision to whole system, without inciting somebody to action It is transferred to other autonomous satellites and carries out decision, and the framework is complicated, needs to communicate between close star.As it can be seen that every Satellite is all in intelligent level I1On, it can mutually cooperate with completion task.In the more star co-architectures of distributing, intelligence is only included Level is in I1Satellite individual, as shown in Figure 1.
Each satellite individual will pass through being circulated throughout for " contexture by self dispatching --- communication interaction --- Coordination Decision " Journey, wherein intelligent level I1Including five main autonomic functions include：
1) programming dispatching
Programming dispatching is divided into two levels：This star and his star.Each star will not only be updated the task scheduling of itself, Also the task scheduling of other satellites is safeguarded.
This star programming dispatching refers to from pending observation mission set, and selection task simultaneously generates observation mission sequence, So that object function maximizes.The task scheduling of itself and resource information are simultaneously sent to Coordination Decision mould by programming dispatching module Block.When the Coordination Decision for receiving Coordination Decision module feedback as a result, being updated to itself task scheduling according to it.
His star programming dispatching refers to the set of tasks sent for Coordination Decision module, his star resource information and punching This star after prominent resolution carries out weight programming dispatching respectively with his star task scheduling, by his star task scheduling after weight programming dispatching Feed back to Coordination Decision module.
2) it executes
It is defended according to the ability of load, state, performance and the obtained task scheduling of programming dispatching module, generation on star Star action sequence and corresponding resource arrangement, and executable action command is sent to each satellite borne equipment.It is complete in task execution At later, image data is sent to data processing module.
3) data processing
For acquired image, target detection, extraction and positioning are carried out, suspicious object point is gone out with therefrom screening, is gone forward side by side One step obtains the multidimensional characteristic (such as orientation angles, shape size etc.) of suspicious object point.If it was found that suspicious object, generates New mission requirements send it to Coordination Decision module.
4) Coordination Decision
When receiving the task execution intention that his star passes over, for identical mission requirements, generated with this star Task execution is intended to, compared to pair, task conflict resolution be carried out according to certain collaborative strategy, and by the task-set after conflict resolution It closes, the task scheduling of his star resource information and this star and his star is sent to programming dispatching module.Programming dispatching module is transmitted His the star task scheduling come is as Coordination Decision as a result, feeding back to his star accordingly.
5) communication interaction
Communication interaction module carries out communication interaction according to communication mechanism seclected time point, and the task of scheduling to be planned is needed Information is sought, as a result, executing intention, resource information and task scheduling etc. transmit away Coordination Decision, and receive his star and pass Pass the packet come.
Information flow between autonomous satellite refers to executing the data needed for task action and disappearing needed for triggering mission requirements Breath.Fig. 2 illustrates the information flow chart in distributing co-architecture, wherein " m " expression message, " d " expression data.Following table is listed The specific descriptions of each message and data in distributing co-architecture.
The detailed description table of message and data in information flow chart
Distributed earth observation satellite system have unique orbital characteristics and intermittence communication characteristic, communication constraint compared with Multi-robot system and multiple no-manned plane system are tightened up, and the communication connectivity in system can not real-time ensuring.Therefore, dynamic In state task environment, in face of the communication constraint of computing capability and time-varying on limited star, it is online to design efficiently applicable more stars Collaborative dispatch mechanism and algorithm obtain balance between system effectiveness and communications cost cost, are a very challenging Problem.
Distributed earth observation satellite system is the satellite set being located on different tracks, and task is to earth surface Each hot spot region be observed imaging, such as corresponding to the hot spot of colcanism or forest fire.Each earth observation is defended Star can swing equipped with single imaging load, image center visual field line.Side-sway activity needs duration limit System, therefore the task requests that observation activity is too near to cannot be realized by same satellite.In addition, satellite is equipped with abnormality detection Equipment finds focus incident, and observation mission request is generated on star.
The online arrival of observation mission request is essentially from three main sources：The task that satellite itself generates in constellation, The emergent upper note of the task and earth station that generate and pass on other satellites in constellation.
Earth observation satellite is on low orbit, and inter-satellite link is not always available.For example, RapidEye is Germany The multispectral earth observation satellite project of business of Brandenburg RapidEye AG companies, includes being defended by what five moonlets formed Star seat.Five RapidEye earth observation satellites are deployed on the same sun-synchronous orbit that height is 630 kilometers, and It is equidistantly placed to ensure consistent image-forming condition and short revisit time interval.These satellites are when on its orbit plane each other Between interval about 19 minutes.It can be seen that between five satellites and direct communication link between star is not present, because of the quantity and height of satellite Degree does not allow the presence of inter-satellite link, therefore we off-line operation and can only cooperate with these satellites in earth station.In the present invention Assuming that there are a limited number of communication relay nodes, it is possible to provide more more communication opportunities than in the past.And these relay nodes can be with It rest on the ground, can also be to be deployed in telecommunication satellite of space etc. such as ground communication station or mobile communicational vehicle.
In the on-line scheduling problem of satellite constellation, there are two types of the task of type, routine observation task and emergent observation are appointed Business.The observation that ground control centre receives user is asked and is arranged it, is then uploaded by satellite-ground link generated Surveillance program.In the present invention, it is known as routine observation task using these as the observation request collected in advance.In contrast, from The observation request reached at random on satellite inside and outside constellation is referred to as emergent observation mission, and unpredictalbe contingency tasks need On-line scheduling is carried out in the case of no ground intervening surface.The corresponding financial value of emergent observation mission is usually above routine observation Task, because this generic task has high value and there is an urgent need for quick responses.For more generally situation, observation mission of meeting an urgent need is past Past reached at random in the form of batch.
From wider more earth observation satellite systematic collaboration angles, first, earth observation satellite once can only be to ground On a target be imaged.Secondly, each task be directed to different observed objects, and be between each other it is independent, each Task can be completed by a satellite.Third, satellite scheduling problem are that task requests quantity is significantly greater than available earth observation The excessive orderability question of number of satellite, therefore build task scheduling for each satellite.For between earth observation satellite and task Effectiveness dependence, there are four kinds of classifications：Without dependence (ND), own schedule plan inside dependence (ID), scheduling Cross-dependence relationship (XD) between plan and complicated dependence (CD).And we the problem of in, there are own schedule meters The inside dependence (ID) drawn, it means that for given satellite, if be able to carry out given task and depend on it The schedule situation of other tasks in the works.
For the unified simplicity of statement, unified definition is carried out to character used in the present invention：
Subscript
i,i′,i″:Low rail LEO (Low Earth Orbit) satellite number, i=1,2 ..., n1
g:Communication relay node serial number, h=1,2 ..., n2
j,j′:Emergent observation mission number, j=1,2 ..., u
k:Routine observation mission number, k=1,2 ..., vi
q:Call duration time window number, q=1,2 ..., mig
o,o′,o″:Task Mission Number, o=1,2 ..., l
Parameter amount
H:Entire scheduling interval
tl:The total duration of entire scheduling interval H
n1:The total quantity of LEO satellite in system
n2:Available communication relay number of nodes
u:Contingency tasks quantity in a collection of subtask
l:The total batch number of contingency tasks in entire scheduling interval H
vi:The normal work to do quantity uploaded on LEO satellite i
ro:Relay node issues the time of o batch contingency tasks to LEO satellite
trio:The time of o batch contingency tasks reaches the time of LEO satellite i
sproll:The side-sway angular speed of LEO satellite
maxθ:The maximum lateral swinging angle of LEO satellite
oesioj:For LEO satellite i, the earliest observation time started of j-th of contingency tasks in o batches
olfioj:For LEO satellite i, the observation end time the latest of j-th of contingency tasks in o batches
olsioj:For LEO satellite i, the observation time started the latest of j-th of contingency tasks in o batches
otwioj:For LEO satellite i, the observation time window of j-th of contingency tasks in o batches
poj:The imaging duration of j-th of contingency tasks in o batches
eoj:The income of j-th of contingency tasks in o batches is provided by decision on manager or star
θioj:For LEO satellite i, the side-sway angle of j-th of contingency tasks in o batches
bioj:For LEO satellite i, the practical time started of j-th of contingency tasks in o batches
cioj:For LEO satellite i, the physical end time of j-th of contingency tasks in o batches
oesik:The earliest start time of upper k-th of the normal work to do of LEO satellite i
olfik:The end time the latest of upper k-th of the normal work to do of LEO satellite i
pik:The imaging duration of upper k-th of the normal work to do of LEO satellite i
eik:The income of upper k-th of the normal work to do of LEO satellite i, is provided by manager
bik:The practical time started of upper k-th of the normal work to do of LEO satellite i
cik:The physical end time of upper k-th of the normal work to do of LEO satellite i
siojk:On LEO satellite i, when j-th of contingency tasks executes in k-th of normal work to do and then o batches Order relies on conversion time
sikoj:On LEO satellite i, when j-th of contingency tasks and then k-th of normal work to do execution in o batches Order relies on conversion time
siojo′j′:On LEO satellite i, when in the o ' batches jth ' a contingency tasks and then in o batches j-th it is emergent Order when task execution relies on conversion time, wherein (o ≠ o ') | | (j ≠ j ')=1
ctwigq:Q-th of call duration time window between LEO satellite i and relay node g
mig:Call duration time number of windows between LEO satellite i and relay node g
utioj:For LEO satellite i, the representation of j-th of contingency tasks in o batches
wioj:When j-th of contingency tasks can dispatch execution on LEO satellite i in o batches, it is equal to 1, otherwise, is equal to 0。
Call duration time window set in communication relay node and system between all LEO satellites, is defined as follows：
Specifically, the call duration time window set between communication relay node g and LEO satellite i, is defined as follows：
Wherein call duration time window ctwigqIt is defined as:
ctwigq=[cesigq,clfigq]
Wherein cesigqIt is earliest communication start time, clfigqIt is sign off time the latest.
The time availability of the time availability ＆ contingency tasks of batch task
When there is one or more call duration time windows between LEO satellite i and relay node, o batches are answered Anxious task can just be distributed to LEO satellite i from relay node.To the availability w of batch taskio, calculate as follows：
Time started ols is being observed the latest and if only if o batches contingency tasksiojLEO satellite i, o crowd are reached before J-th of contingency tasks can be scheduled execution on LEO satellite i in secondary.To the time availability w of contingency tasksioj, calculate such as Under：
Wherein olsioj=max { olfioj-poj-(2·maxθ)/sproll,oesioj}。
According to practical situations, it is as follows that the present invention does setting：
1. the emergent observation mission of separate sources is summarized after node in the communications, and by its publication；
2. load is optical imagery load on star, the influence of ground shadow need to be considered；
Communication delay between 3.LEO satellites and relay node is shorter, can be neglected；
4. it is enough to deposit space and continuous working period admittedly in entire scheduling interval, on star.
When o " batch contingency tasks reach satellite constellation, online cooperative scheduling problem can use mixed integer linear programming Model M ILP (mixed integer linear programming) is indicated：
Decision variable
cioj+(siojk+pik)yiojk+olfioj(yiojk-1)≤cik (4.2)
cik+(sikoj+poj)yikoj+olfik(yikoj-1)≤cioj (4.3)
cio'j'+(sio'j'oj+pioj)yio'j'oj+olfio'j'(yio'j'oj-1)≤cioj (4.4)
(trio+poj)xioj+sikojyikoj+sio'j'ojyio'j'oj≤cioj (4.5)
pikzik+siojkyiojk≤cik (4.6)
(oesioj+poj)xioj≤cioj (4.7)
(oesik+pik)zik≤cik (4.8)
cioj≤olfiojxioj (4.9)
cik≤olfikzik (4.10)
bioj+poj=cio0 (4.11)
bik+pik=cik (4.12)
ci0=0,
zi0=1,
xioj={ 0,1 }, yiojk={ 0,1 }, yikoj={ 0,1 }, yio'j'oj={ 0,1 }, zik={ 0,1 }, wioj={ 0,1 } (4.15)
In concentration-distributed collaboration framework, Coordination Decision function only configures on Centroid, and other nodes can be into Row Distributed Calculation, and corresponding is the task cooperation distribution method based on market mechanism, including auction mechanism.It claps It is to organize to carry out by auctioner to sell, wherein each earth observation satellite is corresponding with task according to the local message of itself calculating Bid amounts, and submitted a tender with this.Each earth observation satellite is bid to notify auctioner, and then auctioner determines again The person of winning of each task.The algorithm mechanism can guarantee generate Lothrus apterus solution because auctioner only select one it is right Ground observation satellite is used as the person of winning.Since the algorithm mechanism is towards concentration-distributed structure/architecture, then calculating of submitting a tender is distributed It carries out.
The present invention is towards decentralized architecture, using distributed algorithm, especially auction algorithm and consistency protocol, to task Carry out Lothrus apterus distribution, the unreliable situation of communication environment suitable for being extremely difficult to global coherency Situation Awareness.For communication The online cotasking of more stars under constraint dispatches this complex combination optimization problem, and towards decentralized architecture, and the present invention proposes Consistency beam algorithm m-CBBA algorithms based on asynchronous communication.
Using the algorithm of synchronous communication to when carrying out communication and having hard and fast rule definition, meanwhile, synchronization policy can be to logical Letter situation is lined up, and can only could be executed under the triggering of specific event, it ensure that algorithm has in program process There is estimated state.In most of iterative algorithms, it is all made of synchronized algorithm.In simultaneous iterative, each module can execute Parallel computation, shares state variable, then waits for next trigger event, to carry out the next iteration process of algorithm.At this During a, the state of each earth observation satellite is ensured.
It is to need to maintain synchronous behavior using the cost price of synchronized algorithm, especially in distributed and distributed algorithm In, to realize synchronous synergetic behavior, then communication is as it can be seen that collaboration synchronization could be carried out while needing to meet collaboration each side.
Consistency beam algorithm CBBA is a distributed auction algorithm, and complicated constraint is handled in a manner of real-time distributed, It is more earth observation satellite multi-task planning problems to going out on missions allocation plan, then in practical applications very efficiently.
Consistency beam algorithm has multiple features, can be used for realizing the scheduling of heterogeneous system cotasking, it is assumed that NtTo wait for point The general assignment number matched, LtFor the maximum task quantity of every beam, NaFor available earth observation satellite quantity, D is the earth observation Earth observation satellite quantity in satellite network, it is specific as follows：
(1) consistency beam algorithm is towards decentralized architecture；
(2) consistency beam algorithm belongs to polynomial time algorithm, then the computation complexity in beam structure stage is O (NtLt)；
(3) consistency beam algorithm can be in max { Nt,LtNaThe interior convergence of D iteration, to which consistent beam algorithm can be well Adapt to the variation of network size and task quantity；
(4) different solution target, earth observation satellite internal model and related constraint can be suitable by designing Score function embodied.If score function meets marginal benefit attenuation characteristic, it is ensured that the feasible solution got well.
For distributed more star systems, to execute consistency beam algorithm, then need every LEO satellite tool there are six information to Amount, it is specific as follows：
1. task beam,1≤n≤|BUio|.The task beam indicates LEO satellite i Successful set of tasks is chosen and dispatched from o batches of contingency tasks, while being added to the time order and function in beam according to each task It is ranked up.The length of current task beam is | BUio|, and the contingency tasks quantity u being less than in a batch, | BUio|≤u； When task beam is empty, BU is usedio=φ and | BUio|=0 indicates.
2. corresponding programmed sequence,1≤n≤|PAio|.In the programmed sequence Task and task Shu Xiangtong, and for indicating that LEO satellite i executes the concrete order of each task in task beam.Programmed sequence Length is identical as the length of task beam, | BUio|=| PAio|≤u。
3. time arrow is executed,1≤n≤|TIio|.It executes time arrow and indicates LEO The practical time started of each task in satellite i executive plan sequences, while vector length is identical as the length of programmed sequence.
4. the person's of winning listLength is u, wherein waiojIndicate that LEO satellite i is currently recognized For the person of winning for j-th of task in o batch contingency tasks, specific value is corresponding with the person's of winning number.Work as waioj= When φ, LEO satellite i thinks currently for the task without the person of winning.
5. acceptance of the bid bid listLength is u, wherein wbiojIndicate the corresponding person of winning to The bid gone out indicates the task currently without the person of winning when value is 0.
6. vector time stampLength is n1, wherein tsioi′It indicates to appoint for o batches are emergent Business, LEO satellite i carry out the newer timestamp of up-to-date information, that is, receive from the places LEO satellite i ' fresh information at the time of point.
Consistency beam algorithm CBBA includes the iteration in two stages：Beam builds the stage and consistency builds the stage, wherein before Person is corresponded to each earth observation satellite and is generated the task beam after sequence in the form of greedy search, and the latter corresponds to identification and appoints Business distribution conflict, and conflict resolution is carried out by the local communication between adjacent earth observation satellite.The two stages are continuous Ground is iteratively repeated, until reaching convergence.
On the basis of consistency beam algorithm CBBA, consistency beam algorithm m-CBBA is improved before original two stages Increase a new stage, synchronous communication loop forecast period.
Since there are stringent communication constraints in practical application, then the strong connectedness between earth observation satellite is impossible Reach, therefore communications loop prediction can take between lifting system overall situation performance and guarantee contingency tasks non-conflict scheduling Must be balanced, rather than realize the global convergence in the person's of winning list.
Stage 1：Synchronous communication loop forecast period
Synchronous communication loop is the bidirectional communication link for two or more earth observation satellites (satellite), with to other Earth observation satellite shared information vector simultaneously receives the feedback information from other earth observation satellites from loop.One same Step communications loop is implemented in the base that there is overlapping time between promoter and the call duration time window of one or more respondents On plinth.
Its description form is multi-component system ＜ ID, I, R, ES, LF a ＞SCL：
- ID is an identifier；
- I is communication initiator number, i.e., first is shared its information；
- R is the communication response person number for receiving simultaneously sharing information；
- ES is the earliest start time of communications loop；
- LF is the end time the latest of communications loop.
Wherein communication initiator I and respondent R is LEO satellite, and ES and LF are by promoter and all respondents Call duration time window on overlapping time section determine.
What it is per subsynchronous shared feedback interaction is all to trigger to be formed on an available synchronous communication loop, wherein sharing Upstream time stamp is identical as downgoing time stamp is shared, and feeding back uplink timestamp is identical as feeding back downlink timestamp.Corresponding Description form ＜ id, i, i ', tUS-DS,tUF-DF＞ is derived from synchronous communication loop ＜ ID, I, R, ES, LF ＞SCL, wherein i ∈ I, i ' ∈ R, tUS-DS,tUF-DF∈ [ES, LF], tUS-DS＜ tUF-DF。
For each LEO satellite, when the emergent observation mission of a batch reaches, be required for predicting itself as promoter can With synchronous communication loop.In addition, when one group of shared information vector reaches, need to synchronize itself in response this logical Letter loop is recorded.By above-mentioned two situations, synchronous communication ring can be established from the angle of itself and be updated to each satellite Road.For each satellite, all synchronous communication loops are updated and are ranked up according to the priority of earliest start time And execution.
In order to realize local local convergence, in the available time interval of loop, whenever loop promoter or respondent institute When the task beam of construction changes, then the subsynchronous shared feedback interaction of triggering one on loop, while generating last time information more New timestamp.
A shared feedback interaction in synchronous communication loop is regarded as once communicating, then in the online association of m-CBBA algorithms Be up to un in scheduling1(n1- 1) secondary communication.In shared feedback interaction every time, following two stages are carried out.
Stage 2：Beam builds the stage
Compared to all possible task beam is enumerated, in m-CBBA, each LEO satellite constructs respective Business beam, and updated with the progress of scheduling process.In this stage of algorithm, each LEO satellite is with succession greediness side Task is constantly added in its task beam by formula, until cannot add any task again.Task in task beam is basis That adds successively is tactic, and the task in executive plan is ranked up according to its corresponding practical time started.
Beam building process is based on the on-line scheduling algorithm towards single star.It is available for being not currently in each of task beam Its financial value is compared by task, LEO satellite with the task score in current acceptance of the bid vector, if bigger, is left new Task get the bid score.For unscheduled set of tasks, earth observation satellite therefrom chooses the task with top score, And be added in task beam, and task beam, executive plan, timestamp, the person of winning and acceptance of the bid vector are updated, with Including newly added task.
The recursive procedure of beam structure is continued until that task beam has reached its capacity limit, or there is no tasks to add It is added in task beam, i.e., the satellite is no longer able to the execution of remaining task to surpass other satellites.
Stage 3：Consistency builds the stage
Once earth observation satellite builds the task beam of itself, then need to be communicated with other earth observation satellites, With resolution distribution conflict.After receiving the person of winning and corresponding acceptance of the bid information that adjacent earth observation satellite passes over, often During a earth observation satellite can determine whether any task in its task beam is won simultaneously by other earth observation satellites Mark.
In the consistency structure stage, each pair of adjacent earth observation satellite can synchronize shared following information vector：The person of winning List WAio, acceptance of the bid bid list WBioWith vector time stamp TSio, corresponding to indicate newest received by other earth observation satellites The timestamp of information update.
For the every message transmitted between sender i and responder i ', earth observation satellite i ' can be according to being received The information arrived executes sequence of operations, to be updated to the information vector of its own.These concrete operations include by it Self information WAi′o, WBi′oAnd TSi′o, it is compared with the information corresponding to earth observation satellite i, to determine for each Which is newest information to task.Earth observation satellite i ' can take three kinds of possible operations to each task j：
1. update:wai′oj=waioj,wbi′oj=wbioj；
2. resetting:wai′oj=φ, wbi′oj=0；
3. giving up:wai′oj=wai′oj,wbi′oj=wbi′oj；.
When being responded to the earth observation satellite i message sent, for contingency tasks oj, earth observation satellite i ' institutes The operation decision rule of use, such as table 1.
Operation decision rule under 1 synchronous communication of table when it receives a message
If the person's of winning list WA in communication processioOr acceptance of the bid bid list WBioIt is changed, then earth observation is defended Star, which checks, to be updated or whether the resetting of the task is in the task beam of itself.If it is present in task beam these tasks and its Subsequent task is required for emptying release, then whenFirst task location serial number won in task beam, then for Subsequent all tasks in task beamAnd its task footnote, it can be updated as follows：
To which task beam is cropped, to remove subsequent task：
Meanwhile corresponding task entry is also removed from executive plan and vector time stamp.
If follow-up scheduled task does not discharge, result is overly conservative, can lead to the reduction of algorithm income, then release is follow-up Scheduled task, re-start beam structure, be to be determined by the surrounding time relevance of task scheduling arrangement.So far, it calculates Method reenters second stage, and new task is added in task beam.M-CBBA algorithms change in latter two stage In generation, is until information variable is no longer formed into columns.In practical applications, the synchronous sharing Times of Feedback on same synchronous communication loop can be by To the limitation of communications cost cost.
Synchronous communication loop prediction computation complexity beGiven one is specific LEO satellites, for the on-line scheduling algorithm of a collection of contingency tasks, computation complexity isThen m- The overall calculation complexity of CBBA algorithms is
Asynchronous communication is applied to the case where being run independently of each other between modules in algorithm.Therefore, asynchronous communication is dividing Extensive use in formula algorithm is dissipated, available information can be utilized at any time, not according to stringent scheduling node.
In m-CBBA algorithms, the beam structure stage can independently and asynchronously execute, because in this stage earth observation Satellite only needs local local message.However, the consistency structure stage of m-CBBA algorithms can be influenced using synchronous communication strategy The responding ability and performance of whole system, because Strategy of Conflict Resolution requires each earth observation satellite to change in each communication The up-to-date information of each adjacent earth observation satellite is known in generation.As it can be seen that this is the method for using global coherency, wherein Each earth observation satellite shares the current state of oneself, and waits the feedback message of all adjacent earth observation satellites to be received (or waiting for the regular hour), subsequently into next algorithm stage.
For real-time system, algorithmic delay compulsory in this way is not realistic, influences the performance table of algorithm Now and rate of convergence, therefore the present invention proposes m-ACBBA algorithms, uses asynchronous communication loop and asynchronous interference resolution rule.
Compared with m-CBBA algorithms, relay node is not only considered as communication node by m-ACBBA, but also using relaying section The effect of point performance synergist.Therefore, the difference is that m-ACBBA uses asynchronous communication between m-CBBA and m-ACBBA Loop is predicted and asynchronous local decision rule.
Stage 1：Asynchronous communication loop forecast period
When its information vector is shared with relaying by LEO satellite immediately after being scheduled to the emergent observation mission of new a batch Node, relay node are responsible for predicting asynchronous communication loop as synergist.
Asynchronous sub- communications loop (ASCLs, Asynchronous sub-communication loops) be by single star from The dispatching cycle of body, the call duration time window between the deadline and relay node and LEO satellite of contingency tasks codetermine 's.Specific description form is multi-component system ＜ ID, I, R, ES, LF a ＞ASCL, it is identical as the description form of synchronous communication loop, But the difference is that I or R is relay node, when a LEO satellite is determined as I, R is then set as relay node.Otherwise, work as I For relay node when, then R is set as LEO satellite.As it can be seen that call duration time window ctwigqIt is equal to asynchronous sub- communications loop ＜ id,i,g,cesigq,clfigq＞ or ＜ id, g, i, cesigq,clfigq＞.
Asynchronous communication loop, also referred to as ACL are writing a Chinese character in simplified form for English Asynchronous communication loop, table It is shown as tuple ＜ an ID, I, R, ESUS,LFUS,ESDS-UF,LFDS-UF,ESDF,LFDF＞ACL, it is independent different by two or three What step communications loop was constituted.
Fig. 5 a indicate two asynchronous sub- communications loop ＜ ID1,I1,R1,ES1,LF1＞ASCLWith ＜ ID2,I2,R2,ES2,LF2 ＞ASCLAbove-mentioned asynchronous communication loop is formed in the two asynchronous sub- communications loops, it is specific as follows there are five kinds of correlativities：
(1) I and R is LEO satellite.ACL promoters I is the promoter I of previous ASCL1, while after ACL respondent R is The respondent R of one ASCL2。
I=I1
R=R2
(2) the respondent R of previous ASCL1With the promoter I of latter ASCL2It is communication relay node.Work as communication relay When node is interconnected, then the corresponding relay nodes of two ASCL can differ.Otherwise, corresponding relay node is same.
(3)ESUSAnd LFUSIndicate that ACL's shares uplink communication time window.
ESUS=ES1
LFUS=min { LF1,LF2}
ESUS≤LFUS
(4)ESDS-UFAnd LFDS-UFIndicate that ACL's shares downlink-feeding back uplink call duration time window.
ESDS-UF=max { ES1,ES2}
LFDS-UF=LF2
ESDS-UF≤LFDS-UF
(5)ESDFAnd LFDFIndicate the feeding back downlink call duration time window of ACL.
ESDF=max { ES1,ES2}
LFDF=LF1
ESDF≤LFDF
The asynchronous communication loop ACL of Fig. 5 b diagrams is made of three asynchronous sub- communications loop ASCL, specifically includes ＜ ID1, I1,R1,ES1,LF1＞ASCL, ＜ ID2,I2,R2,ES2,LF2＞ASCLWith ＜ ID3,I3,R3,ES3,LF3＞ASCL, these three asynchronous sons Communications loop ASCL forms an asynchronous communication loop ACL according to the sequencing of earliest start time.Equally, these three It is specific as follows there are following correlativities in ASCLs：
(1) I and R is LEO satellite.The promoter I of ACL is the promoter I of first ASCL1, meanwhile,
Respondent R is the respondent R of the last one ASCL3。
I=I1
R=R3
The promoter I of (2) first ASCL1With the respondent R of the last one ASCL3It is identical, to constitute a full communication Loop.
I1=R3
The respondent R of (3) first ASCL1With the promoter I of latter two ASCL2And I3It is communication relay node.
When being interconnected between relay node, then corresponding relay node can be different, are the same relayings otherwise Node.
(4)ESUSAnd LFUSIndicate that ACL's shares uplink communication time window.
ESUS=ES1
LFUS=min { LF1,LF2}
ESUS≤LFUS
(5)ESDS-UFAnd LFDS-UFIndicate that ACL's shares downlink-feeding back uplink call duration time window.
ESDS-UF=max { ES1,ES2}
LFDS-UF=LF2
ESDS-UF≤LFDS-UF
(6)ESDFAnd LFDFIndicate the feeding back downlink call duration time window of ACL.
ESDF=max { ES2,ES3}
LFDF=LF3
ESDF≤LFDF
It is asynchronous share feedback interaction be on available asynchronous communication loop triggering generation.It is handed in the asynchronous feedback of sharing In mutually, there are four timestamps altogether, including share uplink, share downlink, feeding back uplink and feeding back downlink.Its description form ＜ id,i,i′,tUS,tDS,tUF,tDF＞ is derived from asynchronous communication loop ＜ ID, I, R, ESUS,LFUS,ESDS-UF,LFDS-UF,ESDF,LFDF ＞ACL, wherein i ∈ I, i ' ∈ R, tUS∈[ESUS,LFUS], tDS,tUF∈[ESDS-UF,LFDS-UF], tDF∈[ESDF,LFDF], tUS ≤tDS, tDS＜ tUF, tUF≤tDF。
For each LEO satellite, in the pot life section of ACL, when the task Shu Fasheng of promoter or respondent become It is when dynamic, then asynchronous to share feedback interaction and executed on loop.Since communication delay and scheduling calculate the occupied time relatively It is short, then it will share downlink and be equal with the timestamp of feeding back uplink, tDS≈tUF。
Primary asynchronous feedback of sharing on ACL is interactive altogether including communicating three times, that is, shares uplink, share in downlink-feedback Row and feeding back downlink.For a batch of contingency tasks, online cooperative scheduling is carried out using m-ACBBA algorithms, is at most needed 3un1(n1- 1) secondary communication.
Every time it is asynchronous share feedback interaction in, beam structure and consistency structure is performed both by.Compared with m-CBBA algorithms, no It is that m-ACBBA algorithms build the stage in consistency with place, using asynchronous local decision rule.
The arrival of message is out of order in asynchronous communications protocol, for example the message that earlier time generates can give birth to than the later time At information evening arrive.Therefore, update of time stamp mechanism of the earth observation satellite based on message receipt times is not in m-CBBA algorithms It is suitble to this asynchronous condition.Then in the configuration of m-ACBBA algorithms, earth observation satellite of winning is indicated with the timestamp in message The middle target practical moment, i.e.,Length is u, wherein tsiojIt indicates for contingency tasks in o batches This is won target generated time on j, LEO satellite i.Rather than the earth observation satellite message employed in m-CBBA algorithms Renewal time, i.e.,Length is n1, wherein tsioi′It indicates to be directed to o batch contingency tasks, LEO satellite I carries out the newer timestamp of up-to-date information.
In m-ACBBA algorithms, using a series of local resolution rules, without global state information.Meanwhile it should Unification processing can be carried out to out of order message and redundancy by clearing up agreement.
Conflict resolution rule in mACBBA algorithms, is not only recipient (respondent) and bids to the person's of winning list and acceptance of the bid List is updated, and is to determine the message for needing to be shared again.
It is the reason of being shared again：First, reducing the communications burden in network, dividing again for redundancy is avoided It enjoys；Second is that handling the confusion of timestamp in asynchronous system.
Following five kinds of possible action can be used in recipient：
1. update ＆ shares again：The information that recipient comes according to sender updates the person's of winning list, acceptance of the bid bid list And time arrow of winning, then share current update again；
2. giving up ＆ again to share：Recipient does not change its information state, is then carried out to the person's of the winning information itself assert Share again；
3. giving up ＆ no longer to share：Recipient does not change the information state of itself, does not share again to it yet, applied to being connect The message received the situation consistent with existing information.
4. resetting ＆ shares again：Recipient is zeroed out the person of winning and value of winning, and is then carried out to the message that original receives Share again, then eliminates confusion；
5. renewal time, ＆ shared again：Recipient is the person of winning, and when it receives a message, generated time stamp is won to it more It is newly current time, to confirm that it oneself is the person of winning oneself to still believe that.
When being responded to the earth observation satellite i message sent, for contingency tasks oj, earth observation satellite i ' institutes The operation decision rule of use, such as table 3.
Local decision rule under 3 asynchronous communication of table when it receives a message
Asynchronous communication loop prediction computation complexity beFor what is given LEO satellites, the on-line scheduling computation complexity for a batch contingency tasks areThen m-ACBBA is calculated The overall calculation complexity of method is
The target of earth observation task is, for example, to monitor to find, detailed survey and tracking forest fire or volcano eruption activity.More Precisely, forest fire and volcano eruption need automatic monitoring to find, positioning and identification.Occur in these emergency events In the case of, we focus on the capability of fast response of raising system, and the global gain by cooperateing with raising system online.
We are alleviated to a certain extent using geostationary orbit (GEO) telecommunication satellite as communication relay node The situation of shortage is communicated, but communication condition is still limited, constraint is still harsh.First, when communication between GEO and LEO Between window it is more much greater than the call duration time window between earth station and LEO satellite；Secondly, GEO and the ground in its coverage area Face station is constantly connected, to ensure that mission requirements upload the real time availability with data down transmission, so with this communication mode, Collaboration comes true more stars online, can support the Disaster Management in higher level.Fig. 6 gives between space segment and ground segment Available signal intelligence.
It is wherein visible：
(1) when overlay areas of the LEO by GEO, the communication between LEO satellite and GEO satellite is available.The communication chain Road is used for through GEO, realizes the emergent observation mission transmitted in both directions between LEO and earth station, and by low-rate data from LEO Descend into earth station (GS).
(2) communication interaction between GEO and GS can be carried out constantly, because GS is located at the overlay area of GEO.It should Communication link is uploaded for observation mission of meeting an urgent need and low-rate data is downloaded.
(3) the two-way communication interaction between GEO is feasible, realizes and transmits information related with task between GEO.
(4) direct communication can not be carried out between LEO satellite.
In application scenarios, 3 LEO satellites are configured on the same track, and top configures 3 GEO satellites under the line, specifically Satellite parametric reduction is shown in Table 3.The time span of emulation is set as 6h.
The setting of the orbit parameter of 4 three LEO satellites of table and three GEO satellites
The characteristics of configuration scene, is as follows：
1. when three LEO satellites on same track to same target is observed imaging on ground when, LEO_A and Time interval between LEO_B is about 12 minutes, and the time interval between LEO_B and LEO_C is also 12 minutes.
The communication coverage area of 2.GEO satellites is the conical area that a cone angle is 7.5 °.
The observation imaging region of 3.LEO satellites is a rectangle.If being observed imaging to same target on same rail, The changing value of imaging angle is about 20 ° between LEO_A and LEO_B, the changing value of imaging angle between same LEO_B and LEO_C Also about 20 °.
The signal intelligence of current distributed networked research center is analyzed, includes the height towards concentration-distributed structure/architecture Signal intelligence between rail satellite, such as signal intelligence between table 4, and the low rail towards decentralized architecture, such as table 5.
Communication constraint situation in 1 current distributed networked research center of table between high low orbit satellite
Communication constraint situation in 6 current distributed networked research center of table between LEO satellite
Compare four kinds of online Co-schedulings towards distributed networked research center, including the contract net association under individual event task Discuss algorithm SI-CNP, contract net protocol the algorithm BA-CNP, improved consistency beam algorithm m-CBBA under batch task and improvement Asynchronous consistency beam algorithm m-ACBBA.
In m-CBBA and m-ACBBA algorithms, what shared feedback was interactive every time shares uplink, shares downlink, feeding back uplink And feeding back downlink timestamp is from the corresponding earliest start time of communications loop and the time zone between the end time the latest respectively Interior random generation.In each complete communications loop, primary shared feedback interaction is at most carried out.
It is total revenue respectively using three performance indicators, scheduled successful contingency tasks ratio and corresponding communication are total Number.This three indexs are described in detail below：
(1) system total revenue refers to all scheduled successful tasks of the distributed networked research center in entire scheduling interval The sum of income.
(2) scheduled successful contingency tasks ratio refers to scheduled successfully emergent observation in distributed networked research center and appoints Percentage shared by being engaged in entering the emergent observation mission that scheduling solves.
(3) total degree is communicated.Communication total degree is the sum of the number of communications of three kinds of one-way communication situations, is LEO respectively It initiates to communicate to GEO, GEO initiates one-way communication between communication and GEO to LEO.
In order to assess the performance of four kinds of online Co-schedulings, we pay close attention to distributed networked research center and are seen to emergent The response that survey task reaches, therefore two and the relevant parameter amount of task, the arrival rate and camera of emergent observation mission refer to To angle and one and the relevant parameter amount of scheduling mechanism, rolling scheduling period, the generation for testing example is very Important.As it can be seen that the two respectively represent Annual distribution characteristic and the space of emergent observation mission with the relevant parameter amount of task Distribution character, and the rolling scheduling period then determines the scale that on-line scheduling calculates.In addition, other all parameter amounts are equal Generation is uniformly distributed by given.With the relevant parameter facilities of satellite, task and scheduling mechanism, as shown in table 7.
2 associated arguments of table are arranged
Fig. 7 gives under the different rolling scheduling periods respectively for the m- not interconnected between GEO satellite with interconnection situation CBBA arithmetic results.Compared with the case where not interconnected between GEO satellite, the GEO satellite network to interconnect helps to realize more High total receipts, complete higher proportion of contingency tasks.Because compared with the case where not interconnecting, interconnect so that call duration time Number of windows increases or time interval extends.Meanwhile needs more communication of interconnecting, the communication between especially GEO are handed over Mutually.
The display of table 8 increases with contingency tasks arrival rate, and the total revenue acquired by each algorithm is also increasing.Emergent In the case of task arrival rate is identical, the spatial distribution range of contingency tasks is wider, and acquired total revenue is lower, scheduling Successful contingency tasks ratio is lower.
When rolling scheduling cycle T is 6min, when the room and time distribution of contingency tasks is identical, m-ACBBA and m- CBBA obtains higher total revenue than SI-CNP and BA-CNP, and the total revenue of m-ACBBA is usually above m-CBBA.And work as When rolling scheduling cycle T is 12min, m-ACBBA obtains highest total revenue in four kinds of algorithms, and obtained by SI-CNP algorithms The income arrived is higher than BA-CNP and m-CBBA algorithms.As it can be seen that for BA-CNP and m-CBBA algorithms, the rolling scheduling period is longer, Total revenue is lower, because BA-CNP algorithms are scheduled for batch task, and corresponds to one emergent each dispatching cycle Business batch, while m-CBBA algorithms at most carry out once sharing instead on the synchronous communication loop within each rolling scheduling period Feedback interaction, then the rolling scheduling period is longer, shares that Times of Feedback is fewer, timely responds to dispatch to influence contingency tasks.
For SI-CNP and m-ACBBA algorithms, the rolling scheduling period is longer, and total revenue is higher, especially m-CBBA algorithms The case where number of communications is reduced, total revenue is still increased, and illustrates that the performance of asynchronous communication strategy is excellent.
8 each algorithm of table total revenue acquired under the conditions of different parameters
Table 9 shows that under identical parameter amount, m-ACBBA dispatches successful contingency tasks ratio most in four kinds of algorithms It is high.On the whole in the way of dispatching successful contingency tasks ratio from high to low, each algorithm is ranked up, is successively： m- ACBBA>m-CBBA>BA-CNP>SI-CNP。
9 each algorithm of table contingency tasks scheduling success rates (%) acquired under the conditions of different parameters
Table 10 is shown to be increased with the arrival rate of contingency tasks, and amount of communications increased, and the traffic of SI-CNP Increasing degree is maximum, because it carries out cooperative scheduling to individual task successively.In these four algorithms, the communication time of m-CBBA Number is minimum, and the number of communications of SI-CNP is most, because the structure condition of synchronous communication loop is stringent, then loop quantity is few, in turn Influence communication interaction number.As the rolling scheduling period increases, the corresponding traffics of SI-CNP greatly increase, and other three kinds The corresponding traffic of algorithm is reduced.
Number of communications of the 10 each algorithm of table under the conditions of different parameters
Table 11 shows that three kinds of algorithms are improved compared to SI-CNP in total revenue, and the income of wherein m-ACBBA increases width Degree is maximum.Equally, table 12 shows that three kinds of algorithms are carried compared to SI-CNP algorithms in the successful ratio of contingency tasks scheduling The increase rate of height, wherein m-ACBBA is maximum.
Table 13 is shown to be increased with the arrival rate of contingency tasks, and average calculation times increase.In identical Parameter Conditions Under, the calculating time needed for these four algorithms is almost the same, and matches with practical computing capability on star.
For table 11 when T is 6min compared with SI-CNP, the income of each algorithm increases (%)
For table 12 when T is 6min compared with SI-CNP, the contingency tasks of each algorithm dispatch successfully increment (%)
Scheduling average calculation times (/s) in a dispatching cycle on 13 1 LEO satellites of table
In conclusion when the communications cost cost in system is high, m-CBBA algorithms can be in system total revenue and communication time Balance is obtained between number, and when the communications cost of system is low, m-ACBBA algorithms are that the high system total revenue of acquisition and height are emergent The optimal selection of task scheduling success rates.
The present invention assists the isomorphism distribution earth observation satellite system towards simple task under stringent communication constraint online It is studied with scheduling problem.More star cooperative scheduling problems towards simple task belong to ID [ST-SR-TA] problem.Cause For that can distribute and dispatch multiple tasks on single star, and due to time window constraint and resource (electricity, admittedly deposit) limitation, needle Single star scheduling to some task, can be influenced by other tasks, then there is the dependence inside operation plan.First For call duration time window and observation time window, the batch availability and available time of contingency tasks are defined, builds every batch of Subproblem MILP models when contingency tasks reach；Then, based on single star on-line scheduling mechanism, respectively for concentration-point Cloth framework and decentralized architecture respectively propose two kinds of online cotasking dispatching algorithms.Specifically, towards concentration-distribution Two kinds of algorithms of formula framework are that the contract net protocol under contract net protocol algorithm SI-CNP and batch task under individual event task is calculated Method BA-CNP；Two kinds of algorithms towards decentralized architecture are to improve consistency beam algorithm m-CBBA and improvement asynchronous consistency beam Algorithm m-ACBBA, the two are to be based on synchronous communication and asynchronous communication respectively.Experiment with computing the result shows that, compared to do not interconnect mutually Logical, interconnecting for relay node can make call duration time window increased and extend, and contribute to system total revenue and meet an urgent need The increase of task scheduling success rates.When the communications cost in system is high, m-CBBA algorithms are in balance profit and required communication It is dominant in quantity.When the communications cost cost in system is high, m-CBBA algorithms can system total revenue and number of communications it Between obtain balance, and when the communications cost of system is low, m-ACBBA algorithms are to obtain high system total revenue and high contingency tasks Dispatch the optimal selection of success rates.
It is last it is to be noted that：The above embodiments are merely illustrative of the technical solutions of the present invention, rather than its limitations. It will be understood by those of ordinary skill in the art that：Can with technical scheme described in the above embodiments is modified, or Equivalent replacement of some of the technical features；These modifications or replacements, it does not separate the essence of the corresponding technical solution The spirit and scope of various embodiments of the present invention technical solution.

Claims (7)

1. a kind of online Synergistic method of more stars based on asynchronous communication, which is characterized in that more stars use decentralized architecture, often Earth observation satellite LEO is autonomous satellite, and the intelligence level having the same in framework, and grade difference is not present, described Autonomous satellite is communicated using geostationary orbit telecommunication satellite as relay node, and geostationary orbit telecommunication satellite The effect of synergist is played, is responsible for predicting asynchronous communication loop, for urgent observation mission of meeting an urgent need, every autonomous satellite All pass through the cyclic process of " contexture by self dispatches --- communication interaction --- Coordination Decision ", wherein every autonomous satellite is not only The task scheduling of itself is updated, also the task scheduling of other autonomous satellites in framework is safeguarded, it is described to be based on The online Synergistic method of more stars of asynchronous communication uses the asynchronous consistency beam algorithm based on asynchronous communication.
2. the online Synergistic method of more stars based on asynchronous communication as described in claim 1, which is characterized in that described based on asynchronous The asynchronous consistency beam algorithm of communication includes：Asynchronous communication loop forecast period, beam structure stage and consistency build the stage, In the asynchronous communication loop forecast period, after earth observation satellite is scheduled new a batch emergent observation mission immediately Its information vector is shared with geostationary orbit telecommunication satellite, the geostationary orbit telecommunication satellite is to asynchronous communication loop ACL predicted,
For each earth observation satellite, in the pot life section of ACL, when the task Shu Fasheng of promoter or respondent become When dynamic, then asynchronous feedback of sharing is executed on ACL and is interacted, and is equal downlink is shared with the timestamp of feeding back uplink, tDS≈ tUF, every time it is asynchronous share feedback interaction in, carry out beam structure the stage and consistency build the stage, wherein it is asynchronous to share feedback Interaction is to trigger to generate on available asynchronous communication loop, shares feedback interaction asynchronous, altogether there are four timestamp, institute It includes sharing uplink to state four timestamps, shares downlink, feeding back uplink and feeding back downlink,
Wherein, asynchronous communication loop ACL is made of two or three independent asynchronous sub- communications loop ASCL, the description shape of ASCL Formula is multi-component system ＜ ID, I, R, ES, LF a ＞ASCL,
- ID is an identifier；
- I is communication initiator number, i.e., first is shared its information；
- R is the communication response person number for receiving simultaneously sharing information；
- ES is the earliest start time of asynchronous sub- communications loop；
- LF is the end time the latest of asynchronous sub- communications loop,
I or R is relay node, and when a LEO is determined as I, R is then set as relay node, otherwise, when I is relay node, then R is set as LEO,
In the case where asynchronous sub- communications loop ASCL has the time interval of overlapping, asynchronous communication loop ACL is by two independences Asynchronous sub- communications loop ASCL constitute, it is different in the case where asynchronous sub- communications loop ASCL does not have the time interval of overlapping Step communications loop ACL is made of three independent asynchronous sub- communications loop ASCL.
3. the online Synergistic method of more stars based on asynchronous communication as claimed in claim 2, which is characterized in that build rank in beam Section, task is constantly added in its task beam by every earth observation satellite in a manner of succession greediness, until that cannot add again Until any task, the task in task beam be according to add successively it is tactic, and the task in task scheduling be by It is ranked up according to its corresponding practical time started, builds the stage in beam, each earth observation satellite structure is respectively corresponding Task beam, and update with the progress of scheduling process, for each available task being not currently in task beam, see over the ground It surveys satellite to be compared its financial value with the task score in current acceptance of the bid vector, if the financial value bigger, retains For new task acceptance of the bid score, for unscheduled set of tasks, earth observation satellite therefrom chooses appointing with top score Business, and be added in task beam, the recursive procedure of beam structure is continued until that task beam has reached its capacity limit, or There is no tasks can be added in task beam, i.e., the earth observation satellite is no longer able to surpass it for the execution of remaining task His earth observation satellite.
4. the online Synergistic method of more stars based on asynchronous communication as claimed in claim 2, which is characterized in that in two asynchronous sons Communications loop ＜ ID1,I1,R1,ES1,LF1＞ASCLWith ＜ ID2,I2,R2,ES2,LF2＞ASCLForm the asynchronous communication loop In the case of, in the two asynchronous sub- communications loops, there are following five kinds of correlativities：
(1) I and R is LEO satellite, and ACL promoters I is the promoter I of previous ASCL1, while ACL respondent R is latter ASCL Respondent R2；
I=I1
R=R2
(2) the respondent R of previous ASCL1With the promoter I of latter ASCL2It is communication relay node；When communication relay node phase When intercommunicated, then the corresponding relay nodes of two ASCL can differ, and otherwise, corresponding relay node is same；
(3)ESUSAnd LFUSIndicate that ACL's shares uplink communication time window；
ESUS=ES1
LFUS=min { LF1,LF2}
ESUS≤LFUS
(4)ESDS-UFAnd LFDS-UFIndicate that ACL's shares downlink-feeding back uplink call duration time window；
ESDS-UF=max { ES1,ES2}
LFDS-UF=LF2
ESDS-UF≤LFDS-UF
(5)ESDFAnd LFDFIndicate the feeding back downlink call duration time window of ACL；
ESDF=max { ES1,ES2}
LFDF=LF1
ESDF≤LFDF。
5. the online Synergistic method of more stars based on asynchronous communication as claimed in claim 2, which is characterized in that in three asynchronous sons Communications loop ＜ ID1,I1,R1,ES1,LF1＞ASCL, ＜ ID2,I2,R2,ES2,LF2＞ASCLWith ＜ ID3,I3,R3,ES3,LF3＞ASCL In the case of forming the asynchronous communication loop, in these three asynchronous sub- communications loops, there are following correlativities：
(1) I and R is LEO satellite, and the promoter I of ACL is the promoter I of first ASCL1, meanwhile, respondent R is last The respondent R of a ASCL3,
I=I1
R=R3
The promoter I of (2) first ASCL1With the respondent R of the last one ASCL3It is identical, to constitute a full communication loop,
I1=R3
The respondent R of (3) first ASCL1With the promoter I of latter two ASCL2And I3It is communication relay node, when relaying saves When being interconnected between point, then corresponding relay node can be different, are the same relay nodes otherwise,
(4)ESUSAnd LFUSIndicate that ACL's shares uplink communication time window,
ESUS=ES1
LFUS=min { LF1,LF2}
ESUS≤LFUS
(5)ESDS-UFAnd LFDS-UFIndicate that ACL's shares downlink-feeding back uplink call duration time window,
ESDS-UF=max { ES1,ES2}
LFDS-UF=LF2
ESDS-UF≤LFDS-UF
(6)ESDFAnd LFDFIndicate the feeding back downlink call duration time window of ACL,
ESDF=max { ES2,ES3}
LFDF=LF3
ESDF≤LFDF。
6. the online Synergistic method of more stars based on asynchronous communication as described in claim 4 or 5, which is characterized in that respondent adopts With one kind in following five kinds action：
1) update ＆ shares again：The information that respondent sends according to sender, update the person's of winning list, acceptance of the bid bid list with And time arrow of winning, then share current update again；
2) give up ＆ again to share：Respondent does not change its information state, is then divided again the person's of the winning information itself assert It enjoys；
3) give up ＆ no longer to share：Respondent does not change the information state of itself, does not share again to it yet, is applied to received The message situation consistent with existing information；
4) resetting ＆ shares again：Respondent is zeroed out the person of winning and value of winning, and the message then received to original is divided again It enjoys, then eliminates confusion；
5) renewal time ＆ shares again：Respondent is the person of winning, and when it receives a message, generated time stamp of winning to it is updated to Current time, to confirm that it oneself is the person of winning oneself to still believe that.
7. the online Synergistic method of more stars based on asynchronous communication as claimed in claim 6, which is characterized in that from as transmission When the message that the earth observation satellite of side is sent is responded, for contingency tasks oj, square earth observation in response is defended Decision rule such as following table is operated used by star,
Local decision rule under asynchronous communication when it receives a message
。
CN201711276877.3A 2017-12-06 2017-12-06 Multi-satellite online collaboration method based on asynchronous communication Active CN108377163B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201711276877.3A CN108377163B (en) 2017-12-06 2017-12-06 Multi-satellite online collaboration method based on asynchronous communication

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201711276877.3A CN108377163B (en) 2017-12-06 2017-12-06 Multi-satellite online collaboration method based on asynchronous communication

Publications (2)

Publication Number Publication Date
CN108377163A true CN108377163A (en) 2018-08-07
CN108377163B CN108377163B (en) 2020-10-23

Family

ID=63015774

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201711276877.3A Active CN108377163B (en) 2017-12-06 2017-12-06 Multi-satellite online collaboration method based on asynchronous communication

Country Status (1)

Country Link
CN (1) CN108377163B (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109039428A (en) * 2018-08-17 2018-12-18 中南大学 Repeater satellite single access antenna based on conflict resolution dispatches stochastic search methods
CN115616902A (en) * 2022-11-07 2023-01-17 中国人民解放军国防科技大学 Cluster spacecraft task allocation method and device based on robust auction algorithm

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6721658B2 (en) * 2001-06-14 2004-04-13 The Johns Hopkins University Integrated navigation and communication system for use in distributed spacecraft systems
JP2008268018A (en) * 2007-04-20 2008-11-06 Mitsubishi Electric Corp Earth observation device
CN104270793A (en) * 2014-09-18 2015-01-07 北京邮电大学 Resource allocation method based on satellite cooperative transmission
CN105049343A (en) * 2015-06-29 2015-11-11 中国科学院遥感与数字地球研究所 Distributed satellite data center efficient and cooperative lightweight high-flux gateway system
CN107244427A (en) * 2017-05-31 2017-10-13 北京空间飞行器总体设计部 A kind of main structure allosteric type satellite platform
CN107332606A (en) * 2017-06-27 2017-11-07 杭州电子科技大学 Based on double sampled LEO system difference space-time OFDM coding methods

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6721658B2 (en) * 2001-06-14 2004-04-13 The Johns Hopkins University Integrated navigation and communication system for use in distributed spacecraft systems
JP2008268018A (en) * 2007-04-20 2008-11-06 Mitsubishi Electric Corp Earth observation device
CN104270793A (en) * 2014-09-18 2015-01-07 北京邮电大学 Resource allocation method based on satellite cooperative transmission
CN105049343A (en) * 2015-06-29 2015-11-11 中国科学院遥感与数字地球研究所 Distributed satellite data center efficient and cooperative lightweight high-flux gateway system
CN107244427A (en) * 2017-05-31 2017-10-13 北京空间飞行器总体设计部 A kind of main structure allosteric type satellite platform
CN107332606A (en) * 2017-06-27 2017-11-07 杭州电子科技大学 Based on double sampled LEO system difference space-time OFDM coding methods

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109039428A (en) * 2018-08-17 2018-12-18 中南大学 Repeater satellite single access antenna based on conflict resolution dispatches stochastic search methods
CN109039428B (en) * 2018-08-17 2020-08-21 中南大学 Relay satellite single-address antenna scheduling random search method based on conflict resolution
CN115616902A (en) * 2022-11-07 2023-01-17 中国人民解放军国防科技大学 Cluster spacecraft task allocation method and device based on robust auction algorithm
CN115616902B (en) * 2022-11-07 2023-03-07 中国人民解放军国防科技大学 Cluster spacecraft task allocation method and device based on robust auction algorithm

Also Published As

Publication number Publication date
CN108377163B (en) 2020-10-23

Similar Documents

Publication Publication Date Title
CN108023637A (en) Isomorphic multi-satellite online collaboration method
CN109088667A (en) Isomorphic multi-satellite online collaboration method for simple tasks
CN106516175B (en) Quick imaging satellite independently transports control system and its operational process
Jacquillat et al. An integrated scheduling and operations approach to airport congestion mitigation
Du et al. A new multi-satellite autonomous mission allocation and planning method
CN108510145B (en) Heterogeneous multi-satellite online collaboration method for downward compound task under communication constraint
US7236861B2 (en) Mission planning system with asynchronous request capability
CN112580906A (en) Satellite remote sensing task planning and ground resource scheduling combined solving method
Evler et al. Airline ground operations: Schedule recovery optimization approach with constrained resources
CN105654220A (en) Multi-satellite combined observation method and system
CN109165782A (en) Civil Aviation Airport ground service support personnel dispatching method and its system
Weigang et al. Intelligent computing methods in air traffic flow management
Yao et al. Task allocation strategies for cooperative task planning of multi-autonomous satellite constellation
Chang et al. Mission planning problem for optical video satellite imaging with variable image duration: A greedy algorithm based on heuristic knowledge
Yang et al. Onboard coordination and scheduling of multiple autonomous satellites in an uncertain environment
CN106742086B (en) Quick imaging satellite independently transports control system
Zheng et al. Onboard mission allocation for multi-satellite system in limited communication environment
Li Online scheduling of distributed Earth observation satellite system under rigid communication constraints
CN109951568A (en) A kind of aviation cluster hybrid multilayer formula alliance construction method improving contract net
CN108377163A (en) Multi-satellite online collaboration method based on asynchronous communication
CN109116817A (en) More spacecraft intelligent management systems and its design method based on Agent technology
CN109377075A (en) A kind of method for scheduling task based on prediction prediction
CN106570614A (en) Onboard autonomous distributed task scheduling method
CN108490976A (en) A kind of scheduling system of unmanned plane cluster
CN115545413A (en) Constraint contract network-based multi-satellite autonomous task planning method

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