CN110400002A - A kind of more star imaging task planing methods - Google Patents

A kind of more star imaging task planing methods Download PDF

Info

Publication number
CN110400002A
CN110400002A CN201910540739.4A CN201910540739A CN110400002A CN 110400002 A CN110400002 A CN 110400002A CN 201910540739 A CN201910540739 A CN 201910540739A CN 110400002 A CN110400002 A CN 110400002A
Authority
CN
China
Prior art keywords
task
node
cluster
satellite
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.)
Granted
Application number
CN201910540739.4A
Other languages
Chinese (zh)
Other versions
CN110400002B (en
Inventor
龙军
钱哲曼
陈仕隆
韩笑冬
徐楠
杨凯飞
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.)
Central South University
China Academy of Space Technology CAST
Original Assignee
Central South University
China Academy of Space Technology CAST
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 Central South University, China Academy of Space Technology CAST filed Critical Central South University
Priority to CN201910540739.4A priority Critical patent/CN110400002B/en
Publication of CN110400002A publication Critical patent/CN110400002A/en
Application granted granted Critical
Publication of CN110400002B publication Critical patent/CN110400002B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/004—Artificial life, i.e. computing arrangements simulating life
    • G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Human Resources & Organizations (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Economics (AREA)
  • Strategic Management (AREA)
  • Entrepreneurship & Innovation (AREA)
  • General Physics & Mathematics (AREA)
  • Game Theory and Decision Science (AREA)
  • Quality & Reliability (AREA)
  • Tourism & Hospitality (AREA)
  • Operations Research (AREA)
  • General Business, Economics & Management (AREA)
  • Marketing (AREA)
  • Development Economics (AREA)
  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Computational Linguistics (AREA)
  • Data Mining & Analysis (AREA)
  • Evolutionary Computation (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Educational Administration (AREA)
  • Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)

Abstract

The invention discloses a kind of more star imaging task planing methods, comprising: establishes task model, all the points goal task is indicated using task model；On the basis of track circle time, using the task in track circle time as the node of cluster graph model, and based on the nonoriented edge between each node in cluster constraint condition building cluster graph model, imaging task cluster graph model is obtained；Satisfaction is clustered into the point target task aggregation of constraint condition as cluster task based on heuristic rule, and calculates the lateral swinging angle of cluster task based on median theorem；The constraint condition and objective function of mission planning are constructed and utilized, mission planning directed acyclic graph model corresponding with imaging task dendrogram model is constructed；Task based access control plans directed acyclic graph model, and carries out mission planning using minimax ant group algorithm, obtains more star imaging task programmes.Under different data scales, the present invention can obtain satisfied mission planning result and with good stability.

Description

A kind of more star imaging task planing methods
Technical field
The present invention relates to imaging satellite mission planning field more particularly to a kind of more star imaging task planing methods.
Background technique
Imaging satellite is the platform equipped with Image-forming instrument, is enclosed and is orbited the earth with specific track, according to user demand Imaging is carried out on a surface target to take pictures；Have the characteristics that lasting imaging time length, wide coverage, limited by national boundaries, it is wide It is general to be applied to the fields such as mapping, military surveillance, environmental protection and territorial survey, obtain the great attention of countries in the world.With The increase of family imaging demand leads to that imaging observation can only be arranged to a subset of set of tasks within task planning horizon, can not The more user demands of satisfaction as much as possible.To solve the problems, such as imaging satellite, supply falls short of demand, and more satellites are launched For earth observation, but rare imaging satellite resource is still abnormal precious in face of a large amount of user's imaging demand.
Summary of the invention
The problem of supply falls short of demand based on current imaging satellite, the present invention provide a kind of more star imaging task planing methods, In The programme for seeking a kind of reasonable science between limited imaging satellite resource and a large amount of user's imaging demand, helps to fill Divide using satellite resource and realize the maximization of user demand satisfaction.
To realize the above-mentioned technical purpose, the present invention adopts the following technical scheme:
A kind of more star imaging task planing methods, comprising the following steps:
Step 1, task model is established, all the points goal task is indicated using task model；
Step 2, all the points goal task is built into imaging task cluster graph model；
Building cluster graph model is carried out on the basis of the single track circle time of satellite；With the point target task in track circle time Respectively as 1 node of respective rail circle time part in cluster graph model, and based on cluster constraint condition building cluster artwork Nonoriented edge in type between each node obtains imaging task cluster graph model；
Step 3, cluster constraint condition will be met and the point target observed can be completed in the same imaging band by satellite Task is polymerized to cluster task in imaging task cluster graph model；
Step 4, the constraint condition and objective function of mission planning are constructed；Utilize the constraint condition and target letter of mission planning Number constructs mission planning directed acyclic graph model corresponding with the imaging task dendrogram model that step 3 obtains；
Step 5, task based access control plans directed acyclic graph model, and carries out mission planning using minimax ant group algorithm；
Step 5.1, in the start node of each track circle time of mission planning directed acyclic graph model, it is respectively provided with ant colony；
Step 5.2, it is directed to each track circle time of mission planning directed acyclic graph model, respectively with current orbit circle time The start node initial position mobile as ant individual, mobile using the terminal node of current orbit circle time as ant individual Final position, and using heuristic information and pheromone concentration as the movement rule of ant individual, make all ants in ant colony Ant individual is moved to final position from initial position, obtains the movement routine of all ant individuals of ant colony；
Wherein, the heuristic information between two neighboring node, by the task quantity between two neighboring node, posture machine The size and difference of longitude of dynamic angle construct to obtain；
Step 5.3, the movement routine selected by most ants is selected from all movement routines of each track circle time, is made Work as the optimal movement routine of time iteration cycle for current orbit circle time；
Step 5.4, the optimal movement routine until updating each track circle time to current iteration cycle；
Step 5.5, the pheromone concentration of mission planning directed acyclic graph model, return step are updated using heuristic information 5.2 enter next iteration cycle；
Step 5.6, when reaching iteration termination condition, using the optimal movement routine of each track circle time as finally more Star imaging task programme.
This programme is cluster task by the point target task aggregation of cluster constraint condition is met, so as to efficiently use The imaging satellite resource of limit completes multiple imaging tasks；Then the task after cluster is appointed using minimax ant group algorithm Business planning, and introduce task quantity, attitude maneuver angle and difference of longitude in minimax ant group algorithm and construct heuristic letter Breath and update pheromone concentration, it is ensured that ant carries out the reasonability of Path selection, and guarantee obtains more preferably mission planning side Case.
Further, the task model established in step 1 is MetaTask, statement are as follows: MetaTask IDSet, TimeWindowMap,Lat,Lon,Priority}；
Wherein, IDSet is the set of several single goal mission numbers；If IDSet only includes a number, the task MetaTask is a point target task；If IDSet includes multiple numbers, which is a cluster task； TimeWindowMap is one using the number SatelliteID of satellite as key, with the SEE time window column between satellite and task Table is the map set of value；Lat indicates the latitude of task MetaTask；Lon indicates the longitude of task MetaTask； Priority indicates the priority of task MetaTask；
Dummy satellite in task model is Satellite, statement are as follows:
Satellite{SatelliteID,AllocatedTasks<MetaTask>,Eccentricity, SemiMajorAxis,Inclination,TrueAnomaly,AscNodeRAAN,Perigee,ConeAngle, AttitudeStabilizationTime,AttitudeSpeed}；
Wherein: SatelliteID indicates the number of satellite, and AllocatedTasks indicates that there are SEE times with the satellite The set of tasks of window, Eccentricity indicate the eccentricity of satellite orbit, and SemiMajorAxis indicates half length of satellite orbit Axis, Inclination indicate the inclination angle of satellite orbit, and TrueAnomaly indicates the true anomaly of satellite orbit, AscNodeRAAN indicates that the right ascension of ascending node of satellite orbit, Perigee indicate the argument of perigee of satellite orbit, ConeAngle Indicating the field angle of satellite remote sensor, AttitudeStabilizationTime indicates the satellite remote sensor attitude stabilization time, AttitudeSpeed indicates satellite remote sensor attitude maneuver angular speed；
SEE time window model in task model is TimeWindow, statement are as follows:
TimeWindow{ID,StartTime,EndTime,RollAngle_S,RollAngle_E}；
Wherein: ID indicates the number of SEE time window, at the beginning of StartTime indicates the SEE time window, EndTime indicates the end time of the SEE time window, and RollAngle_S is indicated corresponding to the SEE time window start times Lateral swinging angle, RollAngle_E indicates lateral swinging angle corresponding to the SEE time window end time.
Further, the calculation method of the lateral swinging angle of the cluster task T are as follows:
Calculate cluster task T in all the points goal task lateral swinging angle value range intersection, using obtained intersection as Lateral swinging angle value range △ θ, obtain cluster task T in each point goal task best lateral swinging angle set B；
It sequentially sorts by size to the best lateral swinging angle in best lateral swinging angle set B, obtains set B ' and set of computations B ' Median M；
Judge median M and cluster the relationship of the lateral swinging angle value range △ θ of task T: if median M is in cluster task In the lateral swinging angle value range △ θ of T, then using median M as the lateral swinging angle of cluster task T；If median M is greater than cluster and appoints The upper bound of the lateral swinging angle value range △ θ of business T will then cluster the upper bound of the lateral swinging angle value range of task T as cluster task T Lateral swinging angle；If the lower bound of lateral swinging angle value range △ θ of the median M less than cluster task T will cluster the side of task T Lateral swinging angle of the lower bound of pivot angle value range as cluster task T.
Cluster task lateral swinging angle calculation method of this programme based on median principle, reduces cluster task image quality Loss.For each of cluster task point target task, there is a best lateral swinging angle, i.e. imaging satellite remote sensor is gone to The position of face point target task, it is best to the observation effect of point target task in the position of the best lateral swinging angle；Satellite remote sensing Device is bigger to the difference of the lateral swinging angle and best lateral swinging angle of point target task, poorer to the observation effect of point target task.So Under the factor for considering image quality, the problem of can seeking cluster task best lateral swinging angle conversion are as follows: in the side-sway of cluster task A lateral swinging angle is taken in the value range of angle so that the absolute value that subtracts each other of the best lateral swinging angle of itself and each point target task it And minimum.And according to median principle it is found that being the middle position of n point with the smallest point of the sum of the distance of n point on number axis Number, it follows that the present invention is based on median principles by the median of the best lateral swinging angle of each point goal task in cluster task, Lateral swinging angle as cluster task, it is possible to reduce loss of the cluster task because of image quality caused by lateral swinging angle value.
Further, the cluster constraint condition include: lateral swinging angle constraint, the constraint of maximum available machine time and free time about Beam,
The lateral swinging angle constraint refers to, if several points goal task t1、t2、t3……、tkIt can cluster and appoint for cluster Be engaged in T, then corresponding lateral swinging angle △ θ1、△θ2、△θ3……、△θkIt needs to meet:
The maximum available machine time constraint refers to, if several points goal task can cluster as cluster task T, Middle any two point target task tl、tkIt is both needed to meet:
max(tek,tel)-min(tsk,tsl)≤MaxDuration；
In formula, tsk、tekRespectively indicate point target task tkSEE time window [tsk,tek] initial time and at the end of Between, tsl、telRespectively indicate point target task tlSEE time window [tsl,tel] initial time and the end time, MaxDuration indicates that remote sensor booting executes the maximum duration limitation of Polaroid observation；
The free time constraint refers to, if two point target task tl、tkIt can cluster as cluster task T, need to expire Foot:
WTkl=max (tel-tek,0)；
In formula, WTklIndicate remote sensor in two point target task tlAnd tkBetween imaging free time,Indicate distant Sensor is in two point target task tlAnd tkBetween the pose adjustment time, DminIndicate the attitude stabilization time of remote sensor； Rolll、RollkRespectively indicate two point target task tlAnd tkLateral swinging angle in same track circle time, v indicate remote sensor appearance State controllable velocity；
This is constrained when several points goal task meets lateral swinging angle constraint, maximum available machine time constraint and free time simultaneously When 3 constraint condition, then the several points goal task meets cluster constraint condition, can be polymerized to cluster task T；
It is constrained when two point target tasks are unsatisfactory for space time, but meets lateral swinging angle constraint and maximum available machine time about Beam and exist between the two tasks and meet multiple point target tasks of cluster condition, then two point target tasks meet Transitivity clusters condition, can be polymerized to cluster task T.
Further, the detailed process of step 2 are as follows:
Step 2.1, there are the point target task-set G of SEE time window in i-th of track circle time for acquisition satellite；
Step 2.2, the value range of the lateral swinging angle of point target task is adjusted；
Step 2.3, using each point target task as 1 node, judge any two in point target task-set G Whether point target task meets cluster constraint condition, if meet if between corresponding two nodes of two point target tasks structure Build nonoriented edge；
Step 2.4, all connected graphs for including in point target task-set G are found out, initiating task Clustering Model CGM is obtained；
Step 2.5, nonoriented edge is not present to any two in each of initiating task Clustering Model CGM connected graph Point target task, judge whether to meet transitivity cluster condition, at two point target tasks corresponding two if meeting Nonoriented edge is constructed between node；
Step 2.6, i=i+1, return step 2.1 are enabled；Until obtain include all rail ring second part imaging task it is poly- Class graph model.
Further, the detailed process of step 3 are as follows:
Step 3.1, a connected graph G in the Task clustering graph model CGM of i-th of track circle time of satellite S is obtained, P is enabled For node set included in connected graph G, cluster result set ClusterRes is initialized；
Step 3.2, if P is sky, 3.9 are gone to step；Otherwise, the maximum node p1 of selection degree from node set P；If degree is most Big node is not unique, randomly chooses a node, and all neighbor nodes of node p1 are added to first set p1_ Neigh；Wherein, the degree of node refers to the quantity for the nonoriented edge being connected with the node, and the neighbor node of node refers to and the node Between with nonoriented edge connection node；
Step 3.3, if first set p1_Neigh is sky, 3.8 are gone to step；Otherwise, it is selected from first set p1_Neigh The node that there are most public neighbours with node p1 is taken, and is added to second set p1_MaxCommNeigh；
Step 3.4, if the second set p1_MaxCommNeigh node that includes is unique, p2=p1_MaxCommNeigh is enabled [0], 7 are gone to step；If the node that second set p1_MaxCommNeigh includes is not unique, 3.5 are gone to step；
Step 3.5, choosing from second set p1_MaxCommNeigh has at least with node p1 without the node of edge closing, And third set p1_MinUnRelatedEdge is added；If the node for including in third set p1_MinUnRelatedEdge is only One, p2=p1_MinUnRelatedEdge [0] is enabled, goes to step 3.7；If including in third set p1_MinUnRelatedEdge Node it is not unique, go to step 3.6；
Step 3.6, the maximum node of priority is chosen from third set p1_MinUnRelatedEdge, if priority Maximum node is not unique, then chooses and be worth the smallest node with side right in the nonoriented edge of node p1 composition, and be set to p2；
Wherein, the side right value of nonoriented edge refers to, the imaging free time between the node at nonoriented edge both ends；
Step 3.7, delete side (p1, p2) without edge closing, node p1, p2 are merged into new node p1, from node set P Middle deletion of node p2 updates first set p1_Neigh, goes to step 3.3；
Step 3.8, node p1 is added in cluster result ClusterRes, goes to step 3.2；
Step 3.9, cluster result ClusterRes is exported, is terminated；
Wherein, the model of cluster result ClusterRes be MSITCR < SatelliteID, ClusterResList < ClusterTask > >, be one is as key, to cluster task list ClusterResList using satellite number SatelliteID The key-value pair set of value；And the model for clustering task is ClusterTask { OrbitID, MetaTaskIDSet, TW {StartTime,EndTime},RollAngle}；
In the model of cluster task, OrbitID is the number of satellite orbit circle time, shows that corresponding cluster task is to belong to Which track circle time of satellite；MetaTaskIDSet is the set for forming the point target mission number of the cluster task, and TW is corresponded to SEE time window of the satellite that cluster task and number are SatelliteID in the OrbitID track circle time, StartTime is the time started, and EndTime is the end time, and RollAngle is lateral swinging angle corresponding to cluster task.
Further, the constraint condition of the mission planning are as follows:
sjk+durj+tranjmk+STi≤smk,
In formula, yjkExpression task tjWhether satellite S is arranged iniK-th of track circle time carries out imaging observation, yjk=1 indicates It arranges, yjk=0 indicates not arrange；KiIndicate satellite SiThe number of total coils to orbit the earth within task planning horizon；sjkIt indicates to appoint Be engaged in tjThe time started is observed in k-th of track circle time of affiliated satellite, tj∈Tik；TikIndicate satellite SiHave in k-th of track circle time The candidate tasks set of time window,n2For with satellite SiIn k-th of track circle time having time window Candidate tasks quantity, and have Tik∈Ti；TiIndicate satellite SiAssigned candidate imaging task polymerization,n1For satellite SiAssigned candidate imaging task quantity, has smkExpression task tmThe time started is observed in k-th of track circle time of affiliated satellite, tm∈Tik；durjExpression task tjObservation hold Continuous duration；tranjmkIndicate appearance required between the task j executed in succession and task m in k-th of track circle time of affiliated satellite State conversion time；STiIndicate satellite SiThe attitude stabilization time；EOiIndicate satellite SiWhen carrying out imaging observation, Imaging remote sensing device Rate of energy dissipation；xjhExpression task thWhether task t can be arranged atjAfter execute imaging observation, xjh=1 indicates energy It is enough, xjh=0 expression can not；ESiIndicate satellite SiRate of energy dissipation when carrying out attitude maneuver；anglejhkIt indicates The task t that k-th of track circle time of satellite executes in successionjWith thBetween attitude maneuver angle；viIndicate satellite SiCarry out posture machine Dynamic speed；EiIndicate satellite SiIn the maximum value of track circle time self-energy consumption；
The objective function of the mission planning are as follows:
In formula, N indicates the imaging observation number of activities of mission planning scheme.
Further, using heuristic information and pheromone concentration as the movement of ant individual described in step 5.2 Rule are as follows: ant individual ant_k present node is task tiWhen, calculate its selection taskAs next node Probability, then the maximum task of select probability is as next node, wherein Probability ptjCalculation formula are as follows:
In formula,Indicate the set for all nodes that ant individual ant_k can reach from present node i, τijIt indicates Pheromone concentration value on from node i to the path node j, ηijIndicate heuristic information corresponding to from node i to the path node j Value, α and β respectively indicate the weight of pheromone concentration and heuristic information, q0For constant value, q be one be greater than 0 less than 1 with Machine number；
The heuristic information ηijCalculation method are as follows:
In formula, TCijIndicate the quantity for the cluster task that node i is included to the path node j, angleijIt indicates from node i To the size of the motor-driven angle of the node j attitude of satellite, Math.abs (Lonj-Loni) indicate the exhausted of node i and the difference of longitude of node j To value；
The calculation method of intermediate quantity ψ are as follows:
Pheromone concentration τijIteration update method are as follows:
In formula, antSize indicates the quantity of ant individual in ant colony, QCIndicate single ant individual in the current iteration period Movement routine heuristic information summation, QLIndicate all ant individuals of ant colony on the optimal mobile road in current iteration period The summation of the heuristic information of diameter, QGOptimal movement routine until indicating ant colony all ant individuals to current iteration cycle The summation of heuristic information, q '0It is a fixed constant value, q ' is one and is greater than 0 random number less than 1.
In the movement rule of ant individual and in pheromone concentration more new strategy, randomness mechanism is introduced, it is ensured that ant A possibility that searching further for solution space makes up the deficiency that minimax ant group algorithm is easily trapped into local optimum, to guarantee Obtain more preferably mission planning scheme.
Further, the limit section of the pheromone concentration is [τmin,τmax], when the pheromone concentration being calculated is super Out when limit section, it is modified as follows:
The program can converge on locally optimal solution to avoid minimax ant group algorithm is premature, to guarantee to obtain more excellent Mission planning scheme.
Further, the point target task refers to the imaging task that satellite remote sensor executes on a surface target.
Beneficial effect
The invention proposes a kind of more star imaging task planing methods, and it is poly- to devise the imaging task based on heuristic rule Class algorithm reduces the uncertainty of Task clustering result；The cluster task lateral swinging angle based on median principle is devised to calculate Method reduces the loss of cluster task image quality.Based on minimax ant colony algorithm for optimization design carry out mission planning scheme More star imaging task planning algorithms of local updating；The task quantity that is introduced in mission planning scheme construction process, attitude maneuver Angle and difference of longitude heuristic information ensure that the reasonability in the moving process of ant position；The chance mechanism introduced in algorithm Effectively avoid the deficiency that ant group algorithm is easily trapped into local optimum.Imaging task cluster and imaging task programme part What is be updated successfully plays the role of increase task satisfaction, reduces Riming time of algorithm.Under different data scales, this hair Bright proposed more star imaging task planning algorithms can obtain satisfied mission planning result and with good stability.
Detailed description of the invention
Fig. 1 is remote sensor side-sway schematic diagram；
Fig. 2 is cluster task lateral swinging angle calculation method flow chart of the invention；
Fig. 3 is task of the present invention, satellite, time window relation schematic diagram；
Fig. 4 is Task clustering graph model schematic diagram of the invention；
Fig. 5 is mission planning directed acyclic graph model schematic of the invention；
Fig. 6 is more star imaging task planning algorithm flow charts of the invention；
Fig. 7 is mission planning scheme local updating schematic diagram of the invention.
Specific embodiment
Elaborate below to the embodiment of the present invention, the present embodiment with the technical scheme is that according to development, The detailed implementation method and specific operation process are given, is further explained explanation to technical solution of the present invention.
The present invention proposes a kind of more star imaging task planing methods, including imaging task clustering method and mission planning side Method；Imaging task cluster converts graph theory group for the imaging task clustering problem by building imaging Task clustering graph model Partition problem, and the imaging task clustering algorithm based on heuristic rule is designed, being responsible for will be by imaging satellite same The point target imaging task that observation is completed in imaging band is polymerized to cluster task；Mission planning has by constructing mission planning To acyclic graph model, converts mission planning problem to and ask in figure the problem of including imaging task quantity most paths, and be based on Minimax ant colony algorithm for optimization design carries out the more star mission planning algorithms of mission planning scheme local updating, is responsible for poly- to task Class result is planned, the critical issue of imaging task time window selection is completed.
One, imaging task clusters
1, constraint condition is clustered
Since imaging satellite resource is limited, to efficiently use limited imaging satellite resource, therefore the present invention will meet it is specific Cluster constraint condition and the point target task aggregation that observation can be completed in the same imaging band by the same imaging satellite To cluster task.In the present invention, point target task refers to imaging task.
In the present invention, cluster constraint condition includes lateral swinging angle constraint and time-constrain；
After the lateral swinging angle refers to satellite remote sensor because being directed toward ground target progress attitude maneuver, ground imageable target is defended The angle formed between star remote sensor and sub-satellite point three；
The sub-satellite point refers to satellite in the subpoint on ground；
As shown in Figure 1, σ indicates the field angle of satellite S；Indicate satellite to task tiExecute observation angle when observation, note The observation angleTask t is executed for satelliteiLateral swinging angle；Imaging effect when Imaging remote sensing device alignment ground target is best, will The observation angle of Imaging remote sensing device at this timeSatellite is denoted as to task tiExecute best observation angle θ when observationi；
As long as ground point goal task tiIn the coverage area of remote sensor, for any lateral swinging angleRemote sensor can carry out imaging observation to it；
Under normal circumstances, for observed image quality the considerations of, needs to be modified lateral swinging angle value range, if repairing Lateral swinging angle value range after just are as follows:
The lateral swinging angle constraint refers to: if several observation missions t1、t2、t3……、tkIt can cluster, must have It sets up.
The time-constrain includes maximum available machine time constraint and imaging free time constraint；
What the time span that the booting of Imaging remote sensing device carries out Polaroid observation was limited by, it is set as MaxDuration, The maximum available machine time constraint refers to: if point target observation mission t1、t2、t3……、tkIt can cluster as cluster task T, must have Formula (2) is set up:
max(tek,tel)-min(tsk,tsl)≤MaxDuration formula (2)；
The imaging free time refers to Imaging remote sensing device two point targets task execution imaging observation mistake continuous to ground Cheng Zhong expends the time on the invalid observation area between two point target tasks；
Set up an office goal task tiAnd tkCorresponding lateral swinging angle is respectively Roll in the same track circle timei、Rollk, then at As remote sensor is in two point target tiAnd tkBetween the pose adjustment time are as follows:
Wherein: v (°/s) indicate remote sensor attitude maneuver speed；
If satellite is to task tk、tlSEE time window in same track circle time is respectively [tsk,tek]、[tsl, tel], and set tsk<tsl, then task tkWith tlBetween imaging free time are as follows:
WTkl=max (tel-tek, 0) and formula (4)；
The imaging free time constraint refers to: being set as the attitude stabilization time of remote sensor being Dmin, then two point targets are appointed Be engaged in tiAnd tkIf can cluster in same track circle time, there must be formula (5) establishment:
In addition, perhaps two point target task tiAnd tjBe unsatisfactory for imaging free time constraint, but meet lateral swinging angle constraint and Maximum available machine time constraint, and it is poly- to meet transitivity there are multiple the meeting cluster constraint of the tasks between the two tasks Class, still can be by point target task tiAnd tjIt is aggregated in a cluster task.
2, relationship modeling
One imaging task can be point target task or aggregated obtained cluster task, may defend with more imagings Star is in relationship of multiple track circle time memories between multiple time windows, three as shown in figure 3, and according to the pass between three System establishes such as drag:
(1) SEE time window model TimeWindow, statement are as follows:
TimeWindow{ID,StartTime,EndTime,RollAngle_S,RollAngle_E}
Wherein: ID indicates the number of SEE time window, at the beginning of StartTime indicates the SEE time window, EndTime indicates the end time of the SEE time window, and RollAngle_S is indicated corresponding to the SEE time window start times Lateral swinging angle, RollAngle_E indicates lateral swinging angle corresponding to the SEE time window end time.
(2) task model MetaTask, statement are as follows:
MetaTask{IDSet,TimeWindowMap,Lat,Lon,Priority}；
Wherein: IDSet is the set of several single goal mission numbers；If IDSet only includes a number, the task MetaTask is a point target task；If IDSet includes multiple numbers, which is a cluster task； TimeWindowMap is one using the number SatelliteID of satellite as key, with the SEE time window column between satellite and task Table is the map set of value；Lat indicates the latitude of task MetaTask；Lon indicates the longitude of task MetaTask； Priority indicates the priority of task MetaTask.
(3) dummy satellite is Satellite, statement are as follows:
Satellite{SatelliteID,AllocatedTasks<MetaTask>,Eccentricity, SemiMajorAxis,Inclination,TrueAnomaly,AscNodeRAAN,Perigee,ConeAngle, AttitudeStabilizationTime,AttitudeSpeed}；
Wherein: SatelliteID indicates the number of satellite, and AllocatedTasks indicates that there are SEE times with the satellite The set of tasks of window, Eccentricity indicate the eccentricity of satellite orbit, and SemiMajorAxis indicates half length of satellite orbit Axis, Inclination indicate the inclination angle of satellite orbit, and TrueAnomaly indicates the true anomaly of satellite orbit, AscNodeRAAN indicates that the right ascension of ascending node of satellite orbit, Perigee indicate the argument of perigee of satellite orbit, ConeAngle Indicating the field angle of satellite remote sensor, AttitudeStabilizationTime indicates the satellite remote sensor attitude stabilization time, AttitudeSpeed indicates satellite remote sensor attitude maneuver angular speed.
3, Task clustering graph model constructs
Point target task is indicated using above-mentioned task model, it then will be by satellite same using cluster constraint condition The point target task that observation is completed in a imaging band is polymerized to cluster task in imaging task cluster graph model, and will gather Generic task is equally indicated using above-mentioned task model.The process of cluster task is wherein polymerized in imaging task cluster graph model Are as follows:
The present invention is using point target task as the node clustered in graph model, on the basis of the single track circle time of satellite, structure Build dendrogram MODEL C lusterGraphModel:List<G, V>, abbreviation CGM, CGM are one and are made of several connected graphs It is connected to set of graphs.G represents a connected graph, and V indicates the node collection (vertex i.e. in graph theory) of connected graph G, the nothing between node It is gradually constructed to side by clustering constraint condition；
Wherein, the specific construction step of imaging task cluster graph model is as described below:
Step A1, the point target task-set G of window between obtaining satellite in the presence of i-th of track circle time；
Step A2 is adjusted the value range of the lateral swinging angle of point target task；
Step A3 judges any two in point target task-set G using each point target task as 1 node Whether point target task meets cluster constraint condition, if meet if between corresponding two nodes of two point target tasks structure Build nonoriented edge；
Step A4 finds out all connected graphs for including in point target task-set G, obtains initiating task Clustering Model CGM；
Nonoriented edge is not present to any two in each of initiating task Clustering Model CGM connected graph in step A5 Point target task, judge whether to meet transitivity cluster condition, at two point target tasks corresponding two if meeting Nonoriented edge is constructed between node；
Step A6 enables i=i+1, return step A1；Until obtain include all rail ring second part imaging task cluster Graph model.
For as shown in Figure 4, the building of Task clustering graph model is further described: in Fig. 4 (a), the height of small rectangle Indicate imaging remote sensor visual field scope, task { 1,2,3 }, task { 5,6,7,8,9 }, task { 11,12 }, task 7,9, 10 }, task { 8,11 } can be polymerized to a cluster task；Wherein, task { 1,3 }, task { 5,7 }, task { 5,8 }, task { 5,9 }, task { 6,8 }, task { 6,9 }, task { 7,9 }, task { 10,9 } are unsatisfactory for imaging free time constraint, but meet and pass Passing property cluster, to can also cluster；Other tasks can not be clustered because being unsatisfactory for cluster constraint condition.If several points target Task can cluster, then have a nonoriented edge to be connected between any two point target task.In summary content constructs figure Task clustering graph model shown in 4 (b), and then a partition problem is converted by point target Task clustering problem.
4, the imaging task clustering algorithm based on heuristic rule
The step of imaging task clustering method proposed by the present invention, is as described below:
Step B1 obtains a connected graph G in the Task clustering graph model CGM of i-th of track circle time of satellite S, enables the P be Node set included in connected graph G initializes cluster result set ClusterRes；
Step B2 goes to step B9 if node set P is sky；Otherwise, the maximum node p of selection degree from node set P1； If it is not unique to spend maximum node, a node is randomly choosed, and by node p1All neighbor nodes are added to first set p1_Neigh；Wherein, the degree of node refers to the quantity for the nonoriented edge being connected with the node, and the neighbor node of node refers to and the section Node with nonoriented edge connection between point；
Step B3, if first set p1_ Neigh is sky, goes to step B8；Otherwise, from first set p1In _ Neigh choose with Node p1Node with most public neighbours, and it is added to second set p1_MaxCommNeigh；
Step B4 enables p if the node that second set p1_MaxCommNeigh includes is unique2=p1_MaxCommNeigh [0], 7 are gone to step；If the node that second set p1_MaxCommNeigh includes is not unique, B5 is gone to step；
Step B5 chooses and node p from second set p1_MaxCommNeigh1With at least without the node of edge closing, and Third set p1_MinUnRelatedEdge is added；If the node for including in third set p1_MinUnRelatedEdge is only One, enable p2=p1_MinUnRelatedEdge [0], goes to step B7；If including in third set p1_MinUnRelatedEdge Node it is not unique, go to step B6；
Understanding without edge closing are as follows: if while e1 using while e2 a vertex as vertex, but another vertex is not the public of e2 Neighbours, then e1 is referred to as e2 without edge closing.
Step B6 chooses the maximum node of priority, if priority is most from third set p1_MinUnRelatedEdge Big node is not unique, then chooses and be worth the smallest node with side right in the nonoriented edge of node p1 composition, and be set to p2；
Step B7 deletes side (p1,p2) without edge closing, by node p1、p2Merge into new node p1, from node set P Deletion of node p2, update first set p1_ Neigh, goes to step B3；
Step B8, by node p1It is added in cluster result ClusterRes, goes to step B2；
Step B9 exports cluster result ClusterRes, terminates.
Wherein, the model of cluster result ClusterRes be MSITCR < SatelliteID, ClusterResList < ClusterTask > >, be one is as key, to cluster task list ClusterResList using satellite number SatelliteID The key-value pair set of value；And the model for clustering task is ClusterTask { OrbitID, MetaTaskIDSet, TW {StartTime,EndTime},RollAngle}；
In the model of cluster task, OrbitID is the number of satellite orbit circle time, shows that corresponding cluster task is to belong to Which track circle time of satellite；MetaTaskIDSet is the set for forming the point target mission number of the cluster task, and TW is corresponded to Time window of the satellite that cluster task and number are SatelliteID in the OrbitID track circle time, StartTime It is the time started, EndTime is the end time, and RollAngle is lateral swinging angle corresponding to cluster task.
To the cluster task that each point goal task clusters, the calculation method of lateral swinging angle is as shown in Fig. 2, if obtain Median cluster task lateral swinging angle value range in, then cluster the lateral swinging angle of task size be equal to the median；Such as The fruit median is greater than the upper bound of cluster task lateral swinging angle value range, then the size for clustering the lateral swinging angle of task is equal to lateral swinging angle The upper bound of value range；Otherwise, the size for clustering the lateral swinging angle of task is equal to the lower bound of lateral swinging angle value range.
Two, mission planning
After above-mentioned first part's imaging task cluster, the several points goal task for meeting cluster constraint condition is gathered It is combined into cluster task, and uniformly indicates cluster task using task model above-mentioned, is then needed to the task (packet after cluster The point target task for including polymerization task and not polymerize) pass through Construct question model to generate mission planning scheme.
Described problem model includes mission planning optimization aim model and mission planning directed acyclic graph model；
The mission planning scheme refers to: the imaging task group of observation can be successively executed by one group of two neighboring imaging task At set；Each of mission planning scheme imaging task includes imaging observation beginning and ending time and satellite remote sensor side-sway Angle information；In addition, mission planning scheme further includes task completeness and the attitude of satellite two evaluation indexes of motor-driven angle summation.
1, Construct question model
Parameter in mission planning optimization aim model is defined, as described below:
According to mission planning target and mission planning model parameter, the present invention devises following mission planning model:
Constraint condition:
sjk+durj+tranjmk+STi≤smkFormula (7)；
Mission planning objective function:
Formula (6) indicates that imaging task execution unique constraints, i.e. any one imaging task at most execute once；Formula (7) it is satellite continuous observation time-constrain, satellite S can be arranged in for twoiObservation is executed in succession in k-th of track circle time Imaging task tj、tm, need to meet this constraint condition；Formula (8) is energy constraint condition, indicates satellite SiIn some track The upper limit of energy consumption no more than energy consumption in circle time；Formula (9) is the optimization object function of imaging task planning problem, table Show maximized task quantity performed；
The mission planning scheme different for two, it is preferential to select attitude maneuver angle if their task satisfaction is equal Spend the lesser mission planning scheme of summation.It is expressed as follows with mathematical way:
N indicates the imaging observation number of activities of mission planning scheme in formula (10)；
The present invention constructs mission planning directed acyclic graph model as shown in Figure 5 to more star imaging task planning problems, Each subgraph Gi corresponds to a track circle time in Fig. 5, and white expression task only exists a time window, and black indicates task There are multiple time windows for beautiful purpose, and Fig. 5 is only depicted in each subgraph Gi and is partially directed toward non-E node from S node Directed edge；
2, more star imaging task planning algorithms based on minimax ant group algorithm
In conjunction with mission planning directed acyclic graph model, the present invention devises a kind of more stars based on minimax ant group algorithm Imaging task planning algorithm (Max-Min Ant Colony Optimization-Local Update, MM-ACO-LU), will be more Star imaging task planning problem is converted into the optimal path problem for meeting objective optimization function in figure of asking, algorithm flow such as Fig. 6 institute Show；
The MM-ACO-LU algorithm includes the local updating and information of mission planning scheme construction, mission planning scheme The big module of update three of element；
The mission planning scheme construction is responsible in the construction each subgraph of directed acyclic graph model by node S to node E's Path, all paths constitute final mission planning result；
The construction of the mission planning scheme passes through the mobile completion of ant position, it is assumed that the current location ant ant_k is Task ti, then ant ant_k selects taskRule as next position is as follows:
In the rule,Indicate the set from the ant ant_k present position i all nodes that can reach, τijIndicate the pheromone concentration value on from node i to the path node j, ηijExpression is opened corresponding to from node i to the path node j The hairdo value of information, α and β respectively indicate the weight of pheromone concentration and heuristic information, q0It is a fixed constant value, q is One is greater than 0 random number less than 1.
The present invention introduces randomness mechanism in the movement rule of ant individual, it is ensured that it is empty that ant searches further for solution Between a possibility that, the deficiency that ant group algorithm is easily trapped into local optimum movement routine is made up, so as to obtain optimal task rule The scheme of drawing.
The meaning of formula (11) are as follows: assuming that the current location ant ant_k is i, a random number q is generated, if q≤q0, Then by [τij]α×[ηij]βValue as ant ant_k choose task tjAs the probability of next node, otherwise using ψ as ant Ant k chooses task tjProbability as next node；Final ant fromP is chosen in settjIt is worth maximum node conduct Next node；
The heuristic information ηijCalculation method it is as follows:
In formula (12), TCijIndicate the quantity for the imaging task that node i is included to the path node j, angleijIt indicates From node i to the size of the motor-driven angle of the node j attitude of satellite, Math.abs (Lonj-Loni) indicate node i and node j longitude Absolute value of the difference.
The present invention is opened by task quantity, attitude maneuver angle and the difference of longitude building introduced between two neighboring node Hairdo information, to guarantee that ant individual selects the reasonable of next mobile node in mission planning directed acyclic graph model Property.
In formula (11), the calculation method of ψ is as follows:
The mission planning scheme generated to mission planning scheme constitution step is responsible in the mission planning scheme local updating It is updated, and only finds the path that length is two between two neighboring task, as shown in fig. 7, carrying out mission planning scheme office Portion updates the mission planning scheme that can be more met objective optimization function, thus when reducing the operation of mission planning scheme Between.
The invention proposes a kind of mission planning scheme local updating methods of task based access control quantity priority rule, and step is such as Under:
Step C1, enabling schedule is track circle time O1Mission planning scheme, i=0, t1=schedule.get (i)；
Step C2 goes to step C7, otherwise, goes to step C3 if i is equal to the task quantity that schedule includes；
Step C3, enabling next is t1 in track circle time O1The descendant node set of imaging observation, t2=are not arranged Schedule.get (i+1), j=0；
Step C4, j=j+1 go to step 5, otherwise, go to step C2 if j is less than the size of next set；
Step C5, t3=next.get (j), if t3 meets formula (8) about after being added to schedule equal to t2, and t3 T3 is then added to the i+1 position of schedule, goes to step C6, otherwise, goes to step C4 by beam；
Step C6, i=i+1, t1==schedule.get (i), t2=schedule.get (i+1), go to step C2；
Step C7 terminates.
The Pheromone update is responsible for all to directed acyclic graph after all ants complete the construction of mission planning scheme Pheromone concentration on path is updated, and the present invention devises a kind of pheromones for making full use of global information and local message Update mode, as follows:
Wherein, antSize indicates the quantity of ant individual in ant colony, QCIndicate single ant individual in the current iteration period Movement routine heuristic information summation, QLIndicate all ant individuals of ant colony on the optimal mobile road in current iteration period The summation of the heuristic information of diameter, QGOptimal movement routine until indicating ant colony all ant individuals to current iteration cycle The summation of heuristic information；q′0It is a fixed constant value, q ' is one and is greater than 0 random number less than 1；
QCCalculation it is as follows:
QGCalculation it is as follows:
QLCalculation it is as follows:
SLThe optimal movement routine that the expression current iteration period obtains, SCThe movement that the expression ant current iteration period obtains Path, SGIndicate the optimal movement routine up to the present obtained, QGIndicate that all ant individuals of ant colony are to current iteration cycle The summation of the heuristic information of optimal movement routine only；ηallIndicate the summation of the heuristic information on all paths, function f () is used to calculate the heuristic information summation of respective paths.
The present invention introduces randomness mechanism in pheromone update strategy, it is ensured that ant searches further for solution space Possibility makes up the deficiency that ant group algorithm is easily trapped into local optimum movement routine, so as to obtain optimal mission planning side Case.
It, will in the renewal process for carrying out pheromones to avoid MM-ACO-LU algorithm is premature from converging on locally optimal solution Pheromones on path are limited to [τmin,τmax], it may be assumed that
Above embodiments are preferred embodiment of the present application, those skilled in the art can also on this basis into The various transformation of row or improvement these transformation or improve this Shen all should belong under the premise of not departing from the application total design Within the scope of please being claimed.

Claims (10)

1. a kind of more star imaging task planing methods, which comprises the following steps:
Step 1, task model is established, all the points goal task is indicated using task model；
Step 2, all the points goal task is built into imaging task cluster graph model；
Building cluster graph model is carried out on the basis of the single track circle time of satellite；With the point target task difference in track circle time As 1 node of respective rail circle time part in cluster graph model, and based in cluster constraint condition building cluster graph model Nonoriented edge between each node obtains imaging task cluster graph model；
Step 3, cluster constraint condition will be met and the point target task observed can be completed in the same imaging band by satellite, Cluster task is polymerized in imaging task cluster graph model；
Step 4, the constraint condition and objective function of mission planning are constructed；Using the constraint condition and objective function of mission planning, Construct mission planning directed acyclic graph model corresponding with the imaging task dendrogram model that step 3 obtains；
Step 5, task based access control plans directed acyclic graph model, and carries out mission planning using minimax ant group algorithm；
Step 5.1, in the start node of each track circle time of mission planning directed acyclic graph model, it is respectively provided with ant colony；
Step 5.2, it is directed to each track circle time of mission planning directed acyclic graph model, respectively with the starting of current orbit circle time The node initial position mobile as ant individual, the mobile termination using the terminal node of current orbit circle time as ant individual Position, and using heuristic information and pheromone concentration as the movement rule of ant individual, make all ants in ant colony Body is moved to final position from initial position, obtains the movement routine of all ant individuals of ant colony；
Wherein, the heuristic information between two neighboring node, by the task quantity between two neighboring node, attitude maneuver angle The size and difference of longitude of degree construct to obtain；
Step 5.3, the movement routine that is selected by most ants is selected from all movement routines of each track circle time, as working as Optimal movement routine of the preceding track circle time when time iteration cycle；
Step 5.4, the optimal movement routine until updating each track circle time to current iteration cycle；
Step 5.5, the pheromone concentration of mission planning directed acyclic graph model, return step 5.2 are updated using heuristic information Into next iteration cycle；
Step 5.6, when reaching iteration termination condition, using the optimal movement routine of each track circle time as final more stars at As mission planning scheme.
2. the method according to claim 1, wherein the task model established in step 1 is MetaTask, statement Are as follows: MetaTask { IDSet, TimeWindowMap, Lat, Lon, Priority }；
Wherein, IDSet is the set of several single goal mission numbers；If IDSet only includes a number, the task MetaTask is a point target task；If IDSet includes multiple numbers, which is a cluster task；
TimeWindowMap is one using the number SatelliteID of satellite as key, with the SEE time between satellite and task Window list is the map set of value；Lat indicates the latitude of task MetaTask；Lon indicates the longitude of task MetaTask； Priority indicates the priority of task MetaTask；
Dummy satellite in task model is Satellite, statement are as follows:
Satellite{SatelliteID,AllocatedTasks<MetaTask>,Eccentricity, SemiMajorAxis,Inclination,TrueAnomaly,AscNodeRAAN,Perigee,ConeAngle, AttitudeStabilizationTime,AttitudeSpeed}；
Wherein: SatelliteID indicates the number of satellite, and AllocatedTasks indicates that there are SEE time windows with the satellite Set of tasks, Eccentricity indicate the eccentricity of satellite orbit, and SemiMajorAxis indicates the semi-major axis of satellite orbit, Inclination indicates the inclination angle of satellite orbit, and TrueAnomaly indicates the true anomaly of satellite orbit, AscNodeRAAN table Show that the right ascension of ascending node of satellite orbit, Perigee indicate that the argument of perigee of satellite orbit, ConeAngle indicate satellite remote sensing The field angle of device, AttitudeStabilizationTime indicate satellite remote sensor attitude stabilization time, AttitudeSpeed Indicate satellite remote sensor attitude maneuver angular speed；
SEE time window model in task model is TimeWindow, statement are as follows:
TimeWindow{ID,StartTime,EndTime,RollAngle_S,RollAngle_E}；
Wherein: ID indicates the number of SEE time window, at the beginning of StartTime indicates the SEE time window, EndTime table Showing the end time of the SEE time window, RollAngle_S indicates lateral swinging angle corresponding to the SEE time window start times, RollAngle_E indicates lateral swinging angle corresponding to the SEE time window end time.
3. according to the method described in claim 2, it is characterized in that, the calculation method of the lateral swinging angle of the cluster task T are as follows:
Calculate the intersection of the lateral swinging angle value range of all the points goal task in cluster task T, the side that obtained intersection is used as Pivot angle value range △ θ obtains the best lateral swinging angle set B of each point goal task in cluster task T；
It sequentially sorts, obtains in set B ' and set of computations B ' by size to the best lateral swinging angle in best lateral swinging angle set B Digit M；
Judge median M and cluster the relationship of the lateral swinging angle value range △ θ of task T: if median M is cluster task T's In lateral swinging angle value range △ θ, then using median M as the lateral swinging angle of cluster task T；If median M is greater than cluster task T Lateral swinging angle value range △ θ the upper bound, then will cluster task T lateral swinging angle value range the upper bound as cluster task T's Lateral swinging angle；If the lower bound of lateral swinging angle value range △ θ of the median M less than cluster task T will cluster the side-sway of task T Lateral swinging angle of the lower bound of angle value range as cluster task T.
4. according to the method described in claim 2, it is characterized in that, the cluster constraint condition includes: lateral swinging angle constraint, maximum Available machine time constraint and free time constraint,
The lateral swinging angle constraint refers to, if several points goal task t1、t2、t3……、tkIt can cluster as cluster task T, Then corresponding lateral swinging angle △ θ1、△θ2、△θ3……、△θkIt needs to meet:
The maximum available machine time constraint refers to, if several points goal task can cluster as cluster task T, wherein appointing Anticipate two point target task tl、tkIt is both needed to meet:
max(tek,tel)-min(tsk,tsl)≤MaxDuration；
In formula, tsk、tekRespectively indicate point target task tkSEE time window [tsk,tek] initial time and the end time, tsl、telRespectively indicate point target task tlSEE time window [tsl,tel] initial time and the end time, MaxDuration indicates that remote sensor booting executes the maximum duration limitation of Polaroid observation；
The free time constraint refers to, if two point target task tl、tkIt can cluster as cluster task T, need to meet:
In formula, WTklIndicate remote sensor in two point target task tlAnd tkBetween imaging free time,Indicate remote sensor In two point target task tlAnd tkBetween the pose adjustment time, DminIndicate the attitude stabilization time of remote sensor；Rolll、 RollkRespectively indicate two point target task tlAnd tkLateral swinging angle in same track circle time, v indicate remote sensor attitude maneuver Speed；
This 3 are constrained when several points goal task meets lateral swinging angle constraint, maximum available machine time constraint and free time simultaneously When constraint condition, then the several points goal task meets cluster constraint condition, can be polymerized to cluster task T；
When two point target tasks are unsatisfactory for space time constraint, but meet lateral swinging angle constraint and the constraint of maximum available machine time, And there are the multiple point target tasks for meeting cluster condition between the two tasks, then two point target tasks meet transmitting Property cluster condition, cluster task T can be polymerized to.
5. according to the method described in claim 4, it is characterized in that, the detailed process of step 2 are as follows:
Step 2.1, there are the point target task-set G of SEE time window in i-th of track circle time for acquisition satellite；
Step 2.2, the value range of the lateral swinging angle of point target task is adjusted；
Step 2.3, using each point target task as 1 node, judge any two point mesh in point target task-set G Whether mark task meets cluster constraint condition, constructs nothing between corresponding two nodes of two point target tasks if meeting Xiang Bian；
Step 2.4, all connected graphs for including in point target task-set G are found out, initiating task Clustering Model CGM is obtained；
Step 2.5, any two in each of initiating task Clustering Model CGM connected graph are not present with the point of nonoriented edge Goal task judges whether to meet transitivity cluster condition, in corresponding two nodes of two point target tasks if meeting Between construct nonoriented edge；
Step 2.6, i=i+1, return step 2.1 are enabled；Until obtain include all rail ring second part imaging task dendrogram Model.
6. according to the method described in claim 5, it is characterized in that, the detailed process of step 3 are as follows:
Step 3.1, a connected graph G in the Task clustering graph model CGM of i-th of track circle time of satellite S is obtained, enables P to connect Node set included in logical figure G, initializes cluster result set ClusterRes；
Step 3.2, if P is sky, 3.9 are gone to step；Otherwise, the maximum node p of selection degree from node set P1；If spending maximum Node is not unique, randomly chooses a node, and by node p1All neighbor nodes are added to first set p1_Neigh；Its In, the degree of node refers to the quantity for the nonoriented edge being connected with the node, and the neighbor node of node refers to and has between the node The node of nonoriented edge connection；
Step 3.3, if first set p1_ Neigh is sky, goes to step 3.8；Otherwise, from first set p1It chooses and saves in _ Neigh Point p1Node with most public neighbours, and it is added to second set p1_MaxCommNeigh；
Step 3.4, if the second set p1_MaxCommNeigh node that includes is unique, p is enabled2=p1_MaxCommNeigh [0], Go to step 7；If the node that second set p1_MaxCommNeigh includes is not unique, 3.5 are gone to step；
Step 3.5, it is chosen and node p from second set p1_MaxCommNeigh1With at least without the node of edge closing, and it is added Third set p1_MinUnRelatedEdge；If the node for including in third set p1_MinUnRelatedEdge is unique, enable p2=p1_MinUnRelatedEdge [0], goes to step 3.7；If the section for including in third set p1_MinUnRelatedEdge Point is not unique, goes to step 3.6；
Step 3.6, the maximum node of priority is chosen from third set p1_MinUnRelatedEdge, if priority is maximum Node it is not unique, then choose and be worth the smallest node with side right in the nonoriented edge of node p1 composition, and be set to p2；
Wherein, the side right value of nonoriented edge refers to, the imaging free time between the node at nonoriented edge both ends；
Step 3.7, side (p is deleted1,p2) without edge closing, by node p1、p2Merge into new node p1, deleted from node set P Node p2, update first set p1_ Neigh, goes to step 3.3；
Step 3.8, by node p1It is added in cluster result ClusterRes, goes to step 3.2；
Step 3.9, cluster result ClusterRes is exported, is terminated；
Wherein, the model of cluster result ClusterRes be MSITCR < SatelliteID, ClusterResList < ClusterTask > >, be one is as key, to cluster task list ClusterResList using satellite number SatelliteID The key-value pair set of value；And the model for clustering task is ClusterTask { OrbitID, MetaTaskIDSet, TW {StartTime,EndTime},RollAngle}；
In the model of cluster task, OrbitID is the number of satellite orbit circle time, shows that corresponding cluster task is to belong to satellite Which track circle time；MetaTaskIDSet is the set for forming the point target mission number of the cluster task, and TW corresponds to cluster SEE time window of the satellite that task and number are SatelliteID in the OrbitID track circle time, StartTime are Time started, EndTime are the end times, and RollAngle is lateral swinging angle corresponding to cluster task.
7. according to the method described in claim 2, it is characterized in that, the constraint condition of the mission planning are as follows:
sjk+durj+tranjmk+STi≤smk,
In formula, yjkExpression task tjWhether satellite S is arranged iniK-th of track circle time carries out imaging observation, yjk=1 indicates to arrange, yjk=0 indicates not arrange；KiIndicate satellite SiThe number of total coils to orbit the earth within task planning horizon；sjkExpression task tjIn K-th of track circle time of affiliated satellite observes the time started, tj∈Tik；TikIndicate satellite SiIn k-th of track circle time having time window The candidate tasks set of mouth,n2For with satellite SiIn the candidate of k-th of track circle time having time window Task quantity, and have Tik∈Ti；TiIndicate satellite SiAssigned candidate imaging task polymerization,n1 For satellite SiAssigned candidate imaging task quantity, hassmkExpression task tmAffiliated K-th of track circle time of satellite observes the time started, tm∈Tik；durjExpression task tjObservation duration；tranjmkIt indicates Posture conversion time required between the task j executed in succession and task m in k-th of track circle time of affiliated satellite；STiExpression is defended Star SiThe attitude stabilization time；EOiIndicate satellite SiWhen carrying out imaging observation, the rate of energy dissipation of Imaging remote sensing device；xjhIt indicates Task thWhether task t can be arranged atjAfter execute imaging observation, xjh=1 indicate can, xjh=0 expression can not；ESi Indicate satellite SiRate of energy dissipation when carrying out attitude maneuver；anglejhkExpression is held in succession in k-th of track circle time of satellite Capable task tjWith thBetween attitude maneuver angle；viIndicate satellite SiCarry out the speed of attitude maneuver；EiIndicate satellite SiIn The maximum value of one track circle time self-energy consumption；
The objective function of the mission planning are as follows:
In formula, N indicates the imaging observation number of activities of mission planning scheme.
8. according to the method described in claim 2, it is characterized in that, utilizing heuristic information and information described in step 5.2 Movement rule of the plain concentration as ant individual are as follows: ant individual ant_k present node is task tiWhen, calculate its selection taskAs the probability of next node, then the maximum task of select probability is as next node, wherein probability Calculation formula are as follows:
In formula,Indicate the set for all nodes that ant individual ant_k can reach from present node i, τijIt indicates from section Pheromone concentration value on point i to the path node j, ηijIndicate heuristic information value corresponding to from node i to the path node j, α The weight of pheromone concentration and heuristic information, q are respectively indicated with β0For constant value, q is one and is greater than 0 random number less than 1；
The heuristic information ηijCalculation method are as follows:
In formula, TCijIndicate the quantity for the cluster task that node i is included to the path node j, angleijIt indicates from node i to section The size of the motor-driven angle of the point j attitude of satellite, Math.abs (Lonj-Loni) indicate node i and node j longitude absolute value of the difference；
The calculation method of intermediate quantity ψ are as follows:
Pheromone concentration τijIteration update method are as follows:
In formula, antSize indicates the quantity of ant individual in ant colony, QCIndicate single ant individual in the shifting in current iteration period The summation of the heuristic information in dynamic path, QLIndicate ant colony all ant individuals in the optimal movement routine in current iteration period The summation of heuristic information, QGThe inspiration of optimal movement routine until indicating all ant individuals to current iteration cycle of ant colony The summation of formula information, q '0It is a fixed constant value, q ' is one and is greater than 0 random number less than 1.
9. according to the method described in claim 8, it is characterized in that, the limit section of the pheromone concentration is [τmin,τmax], When the pheromone concentration being calculated exceeds limit section, it is modified as follows:
10. the method according to claim 1, wherein the point target task refers to satellite remote sensor to ground The imaging task that target executes.
CN201910540739.4A 2019-06-21 2019-06-21 Multi-satellite imaging task planning method Active CN110400002B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201910540739.4A CN110400002B (en) 2019-06-21 2019-06-21 Multi-satellite imaging task planning method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201910540739.4A CN110400002B (en) 2019-06-21 2019-06-21 Multi-satellite imaging task planning method

Publications (2)

Publication Number Publication Date
CN110400002A true CN110400002A (en) 2019-11-01
CN110400002B CN110400002B (en) 2021-10-22

Family

ID=68323303

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201910540739.4A Active CN110400002B (en) 2019-06-21 2019-06-21 Multi-satellite imaging task planning method

Country Status (1)

Country Link
CN (1) CN110400002B (en)

Cited By (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111913788A (en) * 2020-06-19 2020-11-10 合肥工业大学 Task scheduling method and system for imaging satellite
CN111947646A (en) * 2020-08-12 2020-11-17 上海卫星工程研究所 Satellite-borne general description method and system of multi-satellite multi-mode maneuvering imaging model
CN112036459A (en) * 2020-08-24 2020-12-04 中南大学 Satellite task merging method and device based on k-means clustering algorithm and storage medium
CN112082532A (en) * 2020-08-19 2020-12-15 长光卫星技术有限公司 Analysis method for multi-satellite-to-multi-demand planning imaging
CN112085378A (en) * 2020-09-04 2020-12-15 中国平安财产保险股份有限公司 Resource allocation method, device, computer equipment and storage medium
CN112257906A (en) * 2020-09-30 2021-01-22 北京控制工程研究所 Imaging satellite autonomous task planning driving method based on state management
CN112529317A (en) * 2020-12-17 2021-03-19 中国科学院空天信息创新研究院 Satellite imaging task planning method and device, electronic equipment and storage medium
CN113222468A (en) * 2021-06-02 2021-08-06 中国电子科技集团公司第五十四研究所 Imaging satellite resource scheduling method based on deep reinforcement learning
CN113486491A (en) * 2021-05-21 2021-10-08 北京控制工程研究所 Satellite autonomous task planning constraint condition self-perfecting method and system
CN113688560A (en) * 2021-07-13 2021-11-23 中南大学 Multi-satellite single-reconnaissance-target-oriented track maneuvering optimization method
CN115909083A (en) * 2022-10-31 2023-04-04 中国科学院软件研究所 Satellite earth observation discrete interest point clustering planning method and device
CN116205466A (en) * 2023-03-31 2023-06-02 中南大学 Scheduling method and device for observation satellite and terminal equipment
CN116362409A (en) * 2023-04-14 2023-06-30 华东师范大学 Method and system for discovering and tracking shared mobile service trend

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101894367A (en) * 2010-05-26 2010-11-24 中国人民解放军国防科学技术大学 Target clustering method of imaging satellite observation scheduling
US20120029812A1 (en) * 2010-07-29 2012-02-02 King Abdul Aziz City For Science And Technology Method and system for automatically planning and scheduling a remote sensing satellite mission
CN108416493A (en) * 2018-01-29 2018-08-17 南京航空航天大学 A kind of quick imaging satellite mission planning method considering drift angle constraint
CN109002966A (en) * 2018-06-25 2018-12-14 湖南国科轩宇信息科技有限公司 A kind of more star mission planning methods based on K mean cluster

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101894367A (en) * 2010-05-26 2010-11-24 中国人民解放军国防科学技术大学 Target clustering method of imaging satellite observation scheduling
US20120029812A1 (en) * 2010-07-29 2012-02-02 King Abdul Aziz City For Science And Technology Method and system for automatically planning and scheduling a remote sensing satellite mission
CN108416493A (en) * 2018-01-29 2018-08-17 南京航空航天大学 A kind of quick imaging satellite mission planning method considering drift angle constraint
CN109002966A (en) * 2018-06-25 2018-12-14 湖南国科轩宇信息科技有限公司 A kind of more star mission planning methods based on K mean cluster

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
HUAN LI ET AL.: ""Task Allocation of Multi-Target Observation for Microsatellite Cluster"", 《IEEE》 *
许语拉等: ""基于团划分的成像侦察任务聚类方法研究"", 《运筹与管理》 *

Cited By (24)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111913788A (en) * 2020-06-19 2020-11-10 合肥工业大学 Task scheduling method and system for imaging satellite
CN111913788B (en) * 2020-06-19 2022-09-30 合肥工业大学 Task scheduling method and system for imaging satellite
CN111947646A (en) * 2020-08-12 2020-11-17 上海卫星工程研究所 Satellite-borne general description method and system of multi-satellite multi-mode maneuvering imaging model
CN111947646B (en) * 2020-08-12 2022-02-08 上海卫星工程研究所 Satellite-borne general description method and system of multi-satellite multi-mode maneuvering imaging model
CN112082532A (en) * 2020-08-19 2020-12-15 长光卫星技术有限公司 Analysis method for multi-satellite-to-multi-demand planning imaging
CN112036459A (en) * 2020-08-24 2020-12-04 中南大学 Satellite task merging method and device based on k-means clustering algorithm and storage medium
CN112036459B (en) * 2020-08-24 2023-12-22 中南大学 Satellite task merging method, device and storage medium based on k-means clustering algorithm
CN112085378A (en) * 2020-09-04 2020-12-15 中国平安财产保险股份有限公司 Resource allocation method, device, computer equipment and storage medium
CN112085378B (en) * 2020-09-04 2023-02-03 中国平安财产保险股份有限公司 Resource allocation method, device, computer equipment and storage medium
CN112257906B (en) * 2020-09-30 2022-08-12 北京控制工程研究所 Imaging satellite autonomous task planning driving method based on state management
CN112257906A (en) * 2020-09-30 2021-01-22 北京控制工程研究所 Imaging satellite autonomous task planning driving method based on state management
CN112529317A (en) * 2020-12-17 2021-03-19 中国科学院空天信息创新研究院 Satellite imaging task planning method and device, electronic equipment and storage medium
CN113486491A (en) * 2021-05-21 2021-10-08 北京控制工程研究所 Satellite autonomous task planning constraint condition self-perfecting method and system
CN113486491B (en) * 2021-05-21 2023-12-12 北京控制工程研究所 Satellite autonomous mission planning constraint condition self-perfecting method and system
CN113222468B (en) * 2021-06-02 2022-04-08 中国电子科技集团公司第五十四研究所 Imaging satellite resource scheduling method based on deep reinforcement learning
CN113222468A (en) * 2021-06-02 2021-08-06 中国电子科技集团公司第五十四研究所 Imaging satellite resource scheduling method based on deep reinforcement learning
CN113688560A (en) * 2021-07-13 2021-11-23 中南大学 Multi-satellite single-reconnaissance-target-oriented track maneuvering optimization method
CN113688560B (en) * 2021-07-13 2023-09-29 中南大学 Orbit maneuver optimization method for multi-star single reconnaissance target
CN115909083A (en) * 2022-10-31 2023-04-04 中国科学院软件研究所 Satellite earth observation discrete interest point clustering planning method and device
CN115909083B (en) * 2022-10-31 2023-08-08 中国科学院软件研究所 Satellite earth observation discrete interest point clustering planning method and device
CN116205466A (en) * 2023-03-31 2023-06-02 中南大学 Scheduling method and device for observation satellite and terminal equipment
CN116205466B (en) * 2023-03-31 2024-02-13 中南大学 Scheduling method and device for observation satellite and terminal equipment
CN116362409A (en) * 2023-04-14 2023-06-30 华东师范大学 Method and system for discovering and tracking shared mobile service trend
CN116362409B (en) * 2023-04-14 2024-04-02 华东师范大学 Method and system for discovering and tracking shared mobile service trend

Also Published As

Publication number Publication date
CN110400002B (en) 2021-10-22

Similar Documents

Publication Publication Date Title
CN110400002A (en) A kind of more star imaging task planing methods
CN107864007B (en) Multi-satellite multi-ground station resource collaborative allocation management method for regional targets
CN116307535B (en) Multi-star collaborative imaging task planning method based on improved differential evolution algorithm
CN106648852B (en) More star method for scheduling task based on double ant colonies and device
CN109948852B (en) Same-orbit multi-point target imaging task planning method for agile satellite
CN105700555B (en) A kind of multiple no-manned plane collaboratively searching method based on gesture game
CN109933842A (en) A kind of mobile target list star mission planning method based on constraint satisfaction genetic algorithm
Xiaolu et al. Multi satellites scheduling algorithm based on task merging mechanism
CN113487221B (en) Space-sky heterogeneous earth observation resource cooperative scheduling method for dynamic target observation
CN107249167A (en) Indoor comprehensive locating platform and localization method
CN106209210B (en) The Information Network method for managing resource of oriented mission
CN112631717B (en) Asynchronous reinforcement learning-based network service function chain dynamic deployment system and method
Kumar et al. Federated control with hierarchical multi-agent deep reinforcement learning
Petrie et al. Solar polar fields and the 22-year activity cycle: observations and models
CN109165858A (en) Multi-satellite scheduling method for large-area target observation
CN113688560B (en) Orbit maneuver optimization method for multi-star single reconnaissance target
CN109960544A (en) Task parallel scheduling method based on data driving type agile satellite
CN109918731B (en) Satellite task planning simulation analysis method and system based on critical path
CN108416493A (en) A kind of quick imaging satellite mission planning method considering drift angle constraint
CN113313355B (en) Remote sensing satellite earth observation task planning method and device based on genetic algorithm
CN113568426B (en) Satellite cluster collaborative planning method based on multiple satellites and multiple loads
CN104680238A (en) To-be-observed task determination method for multi-satellite synergistic earth observation
CN113485409B (en) Geographic fairness-oriented unmanned aerial vehicle path planning and distribution method and system
CN113188547A (en) Unmanned aerial vehicle path planning method and device, controller and storage medium
CN113051815A (en) Agile imaging satellite task planning method based on independent pointer network

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