CN107622348A - A kind of isomery more AUV system tasks coordination approach under task order constraint - Google Patents

A kind of isomery more AUV system tasks coordination approach under task order constraint Download PDF

Info

Publication number
CN107622348A
CN107622348A CN201710837778.1A CN201710837778A CN107622348A CN 107622348 A CN107622348 A CN 107622348A CN 201710837778 A CN201710837778 A CN 201710837778A CN 107622348 A CN107622348 A CN 107622348A
Authority
CN
China
Prior art keywords
mrow
auv
task
msubsup
msub
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN201710837778.1A
Other languages
Chinese (zh)
Inventor
李娟�
张昆玉
鲁赛
张秉健
刘彩云
袁锐锟
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.)
Harbin Engineering University
Original Assignee
Harbin Engineering University
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 Harbin Engineering University filed Critical Harbin Engineering University
Priority to CN201710837778.1A priority Critical patent/CN107622348A/en
Publication of CN107622348A publication Critical patent/CN107622348A/en
Pending legal-status Critical Current

Links

Landscapes

  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The more AUV system tasks coordination approach of isomery, more AUV systems under being constrained the present invention relates to task order mainly by water surface workbench and have assigning for the AUV of difference in functionality is formed, water surface workbench is responsible for the processing of system data, task is distributed optimization, assignment instructions.The AUV of difference in functionality is mainly responsible for receiving assignment instructions, goes to task marine site to perform different tasks.In the present invention, mainly for some the needing to fix execution sequence of the tasks in practical application, optimized by task distribution of the target to AUV of the most short deadline, obtain preferably task allocative decision.The present invention is optimized repeatedly by the continuous pending task to existing change of real-time, constantly ensures that task allocative decision is optimal in present case.

Description

A kind of isomery more AUV system tasks coordination approach under task order constraint
Technical field
The present invention relates to the task coordinate field of the more AUV systems of isomery, and in particular to different under a kind of task order constraint The more AUV system tasks coordination approach of structure.
Background technology
In general, single AUV can be competent at some simple tasks, and single AUV is difficult some ratios of efficient completion More complicated task, such as：To a variety of marine resources of different zones, the comprehensive exploration task dispatching of topographic(al) reconnaissance.This kind of scale Or in the larger task of complexity, single AUV, no matter from energy angle or from the point of view of surveying ability angle, the power that seems is not From the heart.Then more AUVs systems MAUVS just arise at the historic moment, it is intended to are completed entirely again by the multiple AUV common division of labor with cooperating Miscellaneous task, each of which AUV only need to be performed assigned some subtasks, how may have multiple not Congenerous, these subtasks of the AUV reasonable distributions of different performance, so that the operating efficiency highest of the colony, this problem become It is particularly important.MAUVS apply be there has been proposed it is many it is new have a key issue to be solved, including MAUVS institutional framework, Coordination and coordination mechanism, subsurface communication, data fusion, new A UV platforms etc..Domestic and international many research institutions are in these sides The research from single AUV systems to MAUVS transition has been carried out in face, and its system has some following advantage：1) some tasks are single AUV It can not complete, it is necessary to can be successfully completed by MAUVS；2) height of task is concurrently performed in speed with definitely excellent Gesture；3) robotic team of isomery can form that One function is powerful and comprehensive system；4) knowledge, data sharing will make MAUVS obtains knowledge base and method collection more comprehensively, more healthy and stronger；5) diversity that multiple autonomous underwater vehicle system is formed and tissue Diversity can provide the diversity of the incomparable solution of single underwater robot.
For mathematically, more AUV task coordinates optimization problems are a kind of complicated combinatorial optimization problems, and it belongs to task Distribution and resource optimization category.In the case where AUV types and quantity are consistent, task based access control information distributes one to each AUV Or multiple orderly tasks, while task is completed so that whole AUV system overall efficiencies are optimal.Many at present is all single Generic task coordinates and multiclass task coordinate problem, and research direction has following several classes：Concentrate the method for coordinating control；Distributed AC servo system Method；Economic method domestic and foreign literature proposes many general multi-task planning algorithms, with genetic algorithm, particle The intelligent algorithms such as group's algorithm solve, and disposably complete the task distribution between more AUV.For with mission critical sequence constraint The lifting that cooperation problem performs task ability also with AUV turns into one of future developing trend, it is desirable to which AUV can be realized in detection The function of processing, this requires to realize effective self-cooperation between the more AUV of heterogeneous, improves real-time.
The state of the art of more AUV cooperative systems is that multiple robots of cooperation will complete predetermined task as overall, just Must possess the ability of cooperation decision-making, and cooperate and carry out in a structured manner, the form of this structure there is no final conclusion at present.In addition, Decision process is made up of several different stages, and the communication need of different phase is different.Considering the association of multiple autonomous machines When making, cooperation is typically divided into four-stage：1. understand environment；2. the implementation planned；3. planning selection and task distribution；④ Approve and perform monitoring.In addition to the problems such as such as energy, navigation and corresponding sensor exploitation, multiple autonomous underwater vehicle system To there has been proposed many new challenges, multi-platform control strategy, subsurface communication technology, multi-platform Data fusion technique and more The key technologies such as advanced underwater platform are required for researcher constantly to go to explore.
The content of the invention
Present invention generally provides one kind to be directed in heterogeneous task system, is related to a kind of for the different of execution sequence be present The more AUV task coordination methods of configuration.This method has to minimize coordinated allocation of the total Mission Time as target progress task Stronger real-time.
Characterized in that, specifically comprise the following steps：
(1) communication established between surface station and AUV, surface station are obtained each by radio or wireless network AUV initial position message, surface station obtain AUV accurate positional information, equipment by the data exchange between AUV Type information；
(2) surface station initial work is carried out, mission task downloads initial work, clock initial work；Just After beginningization end-of-job, pass through the mission task relevant information of acquisition, including the coordinate information of each mission mission area, marine site Area information, the time that each mission region performs scanning is calculated in advance；
(3) establish with the most short task coordinate model for object function of more AUV task execution times；
(4) surface station is assisted according to task data information and AUV coordinate information to task using ant group algorithm Tuning, each AUV tasks carrying sequence is obtained, and these optimum results and task marine site information are sent to AUV；
(5) after AUV receives assignment instructions, its mission mission area is independently gone to scan for, and scanned through in the marine site Cheng Hou, AUV are floated, and scanning result is returned into water surface workbench immediately；
(6) surface station once receives AUV scanning result, calculate that each AUV completes that current task needs immediately when Between, pending task, re-optimization is carried out to AUV task with reference to the scanning result newly received and before etc., and will optimization As a result each AUV is passed in real time, distributes follow-up task.
Characterized in that, described step (4) specifically includes：
(4.1) acquisition mission mission area, clearance type AUV initial parameter information, initialization ant group algorithm basic parameter, Time parameter；Mission mission area information includes the position coordinates in the marine site, oceanic area information；Clearance type AUV parameter letter Breath mainly includes：Go out thunder AUV cruising speed V1, initial position message, speed V2 when performing clearance task；Ant group algorithm Basic parameter includes：Ant quantity Num_ant, iterations iter_max, heuristic factor importance factor α, pheromones are important Spend factor β, pheromones volatility coefficient ρ, Pheromone Matrix τ；
(4.2) starting point vector sum terminal vector, the numbering that starting point vector is each clearance type AUV are set, and terminal vector is The numbering in each task object marine site；
(4.3) the state transition probability P of all possible paths is calculated according to starting point vector sum terminal vector；
(4.4) paths are selected from state transition probability P using roulette method, the path as this circulation；
(4.5) more ground zero vector sum emphasis vector, i.e., original starting point are substituted by terminal, and terminal originally is eliminated；
(4.6) (4.3)-(4.5) are repeated, until terminal vector is sky；
(4.7) (4.2)-(4.5) are repeated, realizes and travels through until this generation all ants, record the road that each ant is passed by Footpath；
(4.8) the pheromones τ on per paths is updated；
(4.9) repeat (4.2)-(4.8) step to finish until all generation iteration, record appointing for shortest time in all paths Business distribution information；
The calculation formula of state transition probability P in step described above is as follows：
Heuristic function matrix η is the inverse of the time for the navigation that ant will be on the path in above formula, should be equation below：
Pheromone update principle formula in step described above is as follows：
τij(t+1)=Δ τij(t,t+1)+ρ·τij(t)；
I in above-mentioned each formula, j represent the starting and terminal point numbering in path respectively, and k numbers for ant, and m numbers for AUV, Num_ Ant is ant number, allowedkThe destination set for meeting constraints is represented, α is heuristic factor importance factor, and β is letter Plain importance factor is ceased, η is heuristic function matrix, and ρ is pheromones volatility coefficient, and its span is 0~1, is worth bigger explanation The pheromones volatilized are more, and τ is Pheromone Matrix, and D is the distance matrix of path Origin And Destination, V1To go out thunder type AUV's Velocity vector, TijFor ant k hours underway.
Characterized in that, described step (6) specifically includes：
(6.1) determine whether that new task object occurs, if do not occurred, keep original AUV task coordinate strategies； If there is obtaining the coordinate information and quantity information of new task object, carry out following each step；
(6.2) obtain each AUV working condition and it is expected that complete the time of task completed；When calculating and be current The time difference Delta_Time at quarter；Its Delta_Time=0 if AUV is not carried out task；
(6.3) relevant information of task that statistics is still not carried out, with emerging task object information together as The task data of this real-time optimization；
(6.4) relevant parameter of ant group algorithm is initialized, starting point vector sum terminal vector is set, starting point vector is each AUV numbering, terminal vector are the numbering of each task object；
(6.5) the state transition probability P of all possible paths is calculated according to starting point vector sum terminal vector；
(6.6) paths are selected from state transition probability P using roulette method, the path as this circulation；
(6.7) more ground zero vector sum emphasis vector, i.e., original starting point are substituted by terminal, and terminal originally is eliminated；
(6.8) (6.5)-(6.7) are repeated, until terminal vector is sky；
(6.9) (6.4)-(6.7) are repeated, realizes and travels through until this generation all ants, record the road that each ant is passed by Footpath；
(6.10) the pheromones τ on per paths is updated；
(6.11) repeat (6.4)-(6.10) step to finish until all generation iteration, record the shortest time in all paths Task allocation information；
Above-mentioned state transition probability P calculation formula is as follows：
η is the inverse of the time for the navigation that ant will be on the path in formula, and calculation formula is as follows：
I in above-mentioned formula, j represent the starting and terminal point numbering in path respectively, and k numbers for ant, and m numbers for AUV, allowedkThe destination set for meeting constraints is represented, α is heuristic factor importance factor, and β is pheromones importance factor, η is heuristic function matrix, and ρ is pheromones volatility coefficient, and τ is Pheromone Matrix, and D is the distance matrix of path Origin And Destination, V2For the velocity vector for the thunder type AUV that goes out, Delta_Time is time difference vector, TijFor ant k hours underway.
Characterized in that, the more AUV of described heterogeneous are specifically included：
AUV points are different types, and the AUV of each type configures different detection sensing equipments, and each sensing equipment can For completing different tasks, AUV can also carry a variety of sensing equipments simultaneously.
Characterized in that, described surface station specifically includes：
Surface station has the task optimization distribution function of more AUV systems on working mother boat, and can be by allocation result Sent by radio or wireless network；The heterogeneous AUV of configuration different sensors is mainly responsible for going to each marine site to hold Row task, each AUV can go up floatation surface and carry out data exchange with surface station, and data transfer is not needed between AUV.
Characterized in that, described object function specifically includes：
Object function is as follows：
MinT (m)=Tlength(m)+Tmission(m)+Twait(m)；
M is AUV numbering in formula, and T (m) is m-th AUV general assignment duration, Tlength(m) it is m-th of AUV navigation Time, Tmission(m) it is m-th of AUV task execution time, Twait(m) it is that part AUV may go out before task terminates The time of existing wait state.
Characterized in that, described task coordinate model specifically includes：
Described task coordinate model meets the time minimum under following each constraints；Mission area is different sea Domain, can be 1 independent task, can be also made up of N number of task, perform n-th task necessarily using the implementing result of N-1 tasks as Condition (N >=2), i.e., when the first task is after the completion of the execution in the marine site, the execution of second of task could be started：
Constraints 1 is：Each single item task in task sequence must assure that execution and be only performed once；
X in formulai,mFor task allocation information, i is mission number, and m numbers for AUV, and N is AUV quantity, and constraints 1 meets
Constraints 2 is：That AUV is not used by but completing task can not be present；
Used in formulamFor AUV service condition, constraints 2 is represented by usedm≥xi,m。
Present invention generally provides one kind to be directed in heterogeneous task system, is related to more AUV of task order constraint task association Tune method.This method can carry out process to minimize coordinated allocation of the total Mission Time as target progress task in task In carry out task in time and redistribute, ensure that task allocative decision is optimal, the solution procedure of this method in the current situation It is simple and there is stronger real-time.
Brief description of the drawings
Fig. 1 is the execution step and algorithm flow chart of the present invention；
Fig. 2 is the communication scheme of the more AUV systems of isomery；
Fig. 3 is clearance AUV task distribution condition；
Fig. 4 is that ant group algorithm restrains linearity curve；
Fig. 5 is each AUV tasks carrying timing diagram；
Fig. 6 is each AUV fitting-type figure；
Fig. 7 is that each mission task marine site performs the time that scanning needs；
Fig. 8 is clearance AUV task distribution condition；
Fig. 9 is optimized algorithm simulation result material time point result figure.
Embodiment
The more AUV system tasks coordination approach of isomery under a kind of task based access control sequence constraint of the present invention, with reference to specific Row of having a try be described in detail：
This method mainly comprises the following steps：
(1) communication established between surface station and AUV, surface station are obtained each by radio or wireless network AUV initial position message, prepared for Optimal Decision-making and instruction transmission.In this step surface station can by with Data exchange between AUV obtains AUV accurate positional information, types of equipment information.
(2) initial work；Mainly include surface station initialization, mission task is downloaded, clock initialization.This step Primarily to mission task relevant information is obtained, including the coordinate information of each mission mission area, oceanic area information, use In the time for calculating each mission region execution scanning in advance.
(3) establish with the most short task coordinate model for object function of more AUV task execution times, its object function and be：
MinT (m)=Tlength(m)+Tmission(m)+Twait(m)
M is AUV numbering in formula, and T (m) is m-th AUV general assignment duration, Tlength(m) it is m-th of AUV navigation Time, Tmission(m) it is m-th of AUV task execution time, Twait(m) it is that part AUV may go out before task terminates The time of existing wait state.
This model should be the time minimum in the case where meeting following each constraints：
A, each single item task in task sequence must assure that execution and be only performed once.If xi,mDistribute and believe for task Breath, i are mission number, and m is that AUV is numbered, N for AUV quantity then：
This constraints is represented by
B, it is guarantee preciseness, it is impossible to that AUV is not used by but completing task be present.If usedmFor AUV use Situation, then：
So, this constraints is represented by usedm≥xi,m；
(4) surface station uses ant colony according to task data information and clearance type AUV coordinate information to clearance task Algorithm is coordinated and optimized, and obtains each clearance type AUV tasks carrying sequence, and these optimum results and task marine site are believed Breath is sent to clearance type AUV.Task distribution optimized algorithm to clearance type AUV mainly includes following several steps：
(4.1) acquisition mission mission area, clearance type AUV initial parameter information, initialization ant group algorithm basic parameter, Time parameter；Mission mission area information includes the position coordinates in the marine site, oceanic area information；Clearance type AUV parameter letter Breath mainly includes：Go out thunder AUV cruising speed V1, initial position message, speed V2 when performing clearance task；Ant group algorithm Basic parameter includes：Ant quantity Num_ant, iterations iter_max, heuristic factor importance factor α, pheromones are important Spend factor β, pheromones volatility coefficient ρ, Pheromone Matrix τ；
(4.2) starting point vector sum terminal vector, the numbering that starting point vector is each clearance type AUV are set, and terminal vector is The numbering in each task object marine site；
(4.3) the state transition probability P of all possible paths is calculated according to starting point vector sum terminal vector；
(4.4) paths are selected from state transition probability P using roulette method, the path as this circulation；
(4.5) more ground zero vector sum emphasis vector, i.e., original starting point are substituted by terminal, and terminal originally is eliminated；
(4.6) (4.3)-(4.5) are repeated, until terminal vector is sky；
(4.7) (4.2)-(4.5) are repeated, realizes and travels through until this generation all ants, record the road that each ant is passed by Footpath；
(4.8) the pheromones τ on per paths is updated；
(4.9) (4.2)-(4.8) are repeated until all generation iteration finish, the recording the shortest time in all paths of the task is divided With information.
The calculation formula of state transition probability P in step described above is as follows：
Heuristic function matrix η is the inverse of the time for the navigation that ant will be on the path in above formula, should be equation below：
Pheromone update principle formula in step described above is as follows：
τij(t+1)=Δ τij(t,t+1)+ρ·τij(t)
I in each formula above, j represent the starting and terminal point numbering in path respectively, and k numbers for ant, and m numbers for AUV, Num_ Ant is ant number, allowedkRepresent to meet the destination set of constraints, α is heuristic factor importance factor, β For pheromones importance factor, η is heuristic function matrix, and ρ is pheromones volatility coefficient, and its span is 0~1, and value is bigger Illustrate that volatilized pheromones are more, τ is Pheromone Matrix, and D is the distance matrix of path Origin And Destination, V1For the thunder type that goes out AUV velocity vector, TijFor ant k hours underway.
(5) after clearance type AUV receives assignment instructions, its mission mission area is independently gone to scan for, and in the marine site After the completion of scanning, AUV is floated, and scanning result is returned into water surface workbench immediately.
(6) surface station once receives clearance type AUV scanning result, calculates the thunder type AUV that respectively goes out immediately and completes currently The time that task needs, the pending thunder task of going out such as with reference to the scanning result newly received and before, to the task for the thunder type AUV that goes out Re-optimization is carried out, and optimum results are passed to the thunder type AUV that respectively goes out in real time, distributes follow-up thunder task of going out.The water surface works Stand after scanning result is received, following several steps are mainly included to the task coordination method for the thunder type AUV that goes out：
(6.1) new thunder target appearance of going out is determined whether, if do not occurred, keeps the original thunder AUV task coordinates that go out Strategy；If there is obtaining the coordinate information and quantity information of new thunder target of going out, carry out following each step.
(6.2) obtain each thunder type AUV that goes out working condition and it is expected that complete the time of task completed.Calculate with The time difference Delta_Time at current time.Its Delta_Time=0 if certain thunder type AUV that goes out is not carried out task.
(6.3) relevant information of thunder task of going out that statistics is still not carried out, and emerging thunder target information of going out, together Task data as this real-time optimization.
(6.4) relevant parameter of ant group algorithm is initialized, starting point vector sum terminal vector is set, starting point vector goes out to be each Thunder type AUV numbering, terminal vector are the numbering of each thunder target of going out；
(6.5) the state transition probability P of all possible paths is calculated according to starting point vector sum terminal vector；
(6.6) paths are selected from state transition probability P using roulette method, the path as this circulation；
(6.7) more ground zero vector sum emphasis vector, i.e., original starting point are substituted by terminal, and terminal originally is eliminated；
(6.8) (6.5)-(6.7) are repeated, until terminal vector is sky；
(6.9) (6.4)-(6.7) are repeated, realizes and travels through until this generation all ants, record the road that each ant is passed by Footpath；
(6.10) the pheromones τ on per paths is updated；
(6.11) (6.4)-(6.10) are repeated to finish until all generation iteration, records the task of shortest time in all paths Distribute information.
The calculation formula of state transition probability P in step described above is as follows：
η is the inverse of the time for the navigation that ant will be on the path in formula, and calculation formula is as follows：
I in formula above, j represent the starting and terminal point numbering in path respectively, and k numbers for ant, and m numbers for AUV, allowedkThe destination set for meeting constraints is represented, α is heuristic factor importance factor, and β is pheromones importance Factor, η are heuristic function matrix, and ρ is pheromones volatility coefficient, and τ is Pheromone Matrix, and D is the distance of path Origin And Destination Matrix, V2For the velocity vector for the thunder type AUV that goes out, Delta_Time is time difference vector, TijFor ant k hours underway.
The area factor in marine site with specific aim, i.e., is not considered because thunder task of going out compares, and its task execution time is main It is relevant with the number of target.So task time equation below for the thunder type AUV that goes out：
T=N*Td
In formula, N represents the quantity of the pending target in the marine site, TdRepresent go out thunder type AUV processing unit target time.
In real-time optimization procedure, the thunder type that each goes out AUV execution state is probably different, can not ensure these AUV Assign instruction after go to immediately, so must account for go out thunder type AUV complete current task also need to time delta _ Time。
It will further illustrate that the present invention is specific real with 10 marine sites, exemplified by 3 clearance types AUV, 2 thunder type AUV that go out below Mode is applied, the present invention will be described in detail for control accompanying drawing.
(1) communication established between surface station and AUV, surface station are obtained each by radio or wireless network AUV initial position message, prepared for Optimal Decision-making and instruction transmission.In this step surface station can by with Data exchange between AUV obtains AUV accurate positional information, types of equipment information.Wherein surface station obtains 5 AUV Shown in types of equipment Fig. 6.
AUV velocity information is V=[1 1/2 1 1/2 1]；
(2) initial work；Mainly include surface station initialization, mission task is downloaded, clock initialization.This step Primarily to mission task relevant information is obtained, including the coordinate information of each mission mission area, oceanic area information, use In the time for calculating each mission region execution scanning in advance.Wherein in order to embody the universality of algorithm, ten missions are generated at random The three-dimensional coordinate of task, and configure the area in each marine site.Can be pre- according to the sweep speed of oceanic area and clearance type AUV Calculate shown in the total scanning time Fig. 7 in each marine site：
(3) surface station uses ant colony according to task data information and clearance type AUV coordinate information to clearance task Algorithm is coordinated and optimized, and obtains each clearance type AUV tasks carrying sequence, and these optimum results and task marine site are believed Breath is sent to clearance type AUV.Task distribution optimized algorithm to clearance type AUV mainly includes following several steps：
(3.1) task marine site, clearance type AUV detail parameters information are obtained, initializes ant group algorithm basic parameter, time Parameter；Task marine site information includes the position in marine site, area information；Clearance type AUV parameter information mainly includes：Go out thunder AUV Cruising speed, initial position message, perform clearance task when speed；The basic parameter of ant group algorithm includes：Ant quantity Num_ant=30, iterations iter_max=300, heuristic factor importance factor α=1, pheromones importance factor β= 1, pheromones volatility coefficient ρ=0.1, Pheromone Matrix τ is unit matrix；
(3.2) starting point vector start and terminal vector end, the numbering that starting point vector is each clearance AUV, terminal are set Vector is the numbering in each target marine site；Originate starting point vector start=[11,13,15] and (understand clearance type AUV's according to table 1 Numbering is 11,13,15), (it is respectively 1 to set task marine site numbering to starting terminal vector end=[1,2,3,4,5,6,7,8,9,10] ~10)；
(3.3) the state transition probability P of all possible paths is calculated according to starting point vector sum terminal vector；
(3.4) paths are selected from state transition probability P using roulette method, as this circulation path, Assuming that [11,3] this routing information is selected for the first time；
(3.5) more ground zero vector sum emphasis vector, i.e., original starting point are substituted by terminal, and terminal originally is eliminated, Starting point vector start=[3,13,15] after renewal, terminal vector end=[1,2,4,5,6,7,8,9,10]；
(3.6) (3.3)-(3.5) are repeated, until terminal vector is empty, i.e. end=[]；
(3.7) (3.2)-(3.5) step is repeated, realizes and travels through until this generation all ants, record each ant and pass by Path；
(3.8) the pheromones τ on per paths is updated；
(3.9) repeat (3.2)-(3.8) step to finish until all generation iteration, record appointing for shortest time in all paths Business distribution information.
By using the task optimization of ant group algorithm, the algorithmic statement curve by the curve as shown in figure 4, be known that Ant group algorithm is held back by withholding, and algorithm is effective.The clearance type AUV task allocative decisions shown in Fig. 8 can be obtained by optimum results, Detailed path information such as Fig. 3, such as AUV11 go to marine site C2, C1 and C3 to carry out cover type scanning successively.
(4) after clearance type AUV receives assignment instructions, its task marine site is independently gone to carry out covering search, and in the marine site After the completion of scanning, AUV is floated, and scanning result is uploaded into surface station.
(5) after surface station receives clearance AUV scanning result, the thunder type AUV that respectively goes out is calculated immediately and completes current task The time needed, with reference to the scanning result that newly receives and etc. pending thunder task of going out, the task for the thunder type AUV that goes out is carried out again Optimization, and optimum results are passed to the thunder type AUV that respectively goes out in real time, guided for their ensuing thunder tasks of going out.Water surface work Stand after scanning result is received, following several steps are mainly included to the task coordination method for the thunder type AUV that goes out：
(5.1) new thunder target appearance of going out is determined whether, if do not occurred, keeps original thunder type AUV tasks association of going out Adjust strategy；If there is obtaining the coordinate information and quantity information of new thunder target of going out, carry out following each step.
(5.2) obtain each thunder type AUV that goes out working condition and it is expected that complete the time of task completed.Calculate with The time difference Delta_Time at current time.If certain thunder type AUV not its Delta_Time=0 if execution task that goes out.At this A typical time is taken to be analyzed as follows in example：As shown in table 4 below in moment 1277.5s, the thunder AUV14 that now goes out is just in C10 marine sites Execution is gone out thunder task, but new task now occurs, i.e., C5 marine sites find 1 target, and task coordinate optimization at this moment is just Need to consider that AUV14 completes time delta _ Time that current task also needs to.This time is only an estimation, it is impossible to is protected Demonstrate,prove its accuracy.In the real-time optimization at the moment, two AUV Delta_Time is [0,1013.6s] respectively.
(5.3) relevant information of thunder task of going out that statistics is still not carried out, and emerging thunder target information of going out, together Task data as this real-time optimization.
The Optimization Steps of the ensuing task distribution ant group algorithm to the thunder AUV that goes out, it is excellent with clearance AUV task coordinate Change similar, difference is mainly reflected in analysis and consideration to Delta_Time, i.e., the thunder that each goes out is in calculating task total time When need to add this Delta_Time, in this, as the basis of Pheromone update.It will not be described here.
Shown in optimized algorithm simulation result material time point Fig. 9, each AUV execution period is as shown in Figure 5：
The period of each AUV execution tasks shown in Fig. 5 can be seen that to be scanned in clearance AUV to marine site During, the thunder AUV that goes out has begun to handle the target point having now found that.For a piece of marine site for having thunder, two kinds of AUV Task order exist successively, but in whole process the execution time of two task types exist and intersect, this is relative to first with sweeping All marine sites, all search finishes thunder AUV, then goes to the go out situation of thunder of thunder marine site more to save time loss.
In summary, by using the strategy to all task re-optimizations distribution to be done, it can be seen that the present invention Method realizes the purpose of the optimum allocation to known target, improves clearance type AUV and the thunder type AUV that goes out time registration, Shorten the total duration of overall task.

Claims (7)

  1. It is 1. a kind of for the more AUV task coordination methods of heterogeneous of execution sequence be present, it is characterised in that specifically include as Lower step：
    (1) communication established between surface station and AUV, surface station obtain each AUV by radio or wireless network Initial position message, surface station obtains AUV accurate positional information, equipment class by data exchange between AUV Type information；
    (2) surface station initial work is carried out, mission task downloads initial work, clock initial work；Initialization After end-of-job, pass through the mission task relevant information of acquisition, including the coordinate information of each mission mission area, oceanic area Information, the time that each mission region performs scanning is calculated in advance；
    (3) establish with the most short task coordinate model for object function of more AUV task execution times；
    (4) surface station using ant group algorithm coordinate excellent according to task data information and AUV coordinate information to task Change, obtain each AUV tasks carrying sequence, and these optimum results and task marine site information are sent to AUV；
    (5) after AUV receives assignment instructions, its mission mission area is independently gone to scan for, and after the completion of the scanning of the marine site, AUV is floated, and scanning result is returned into water surface workbench immediately；
    (6) surface station once receives AUV scanning result, calculates each AUV immediately and completes the time that current task needs, knot The scanning result that newly receives and the pending task such as before are closed, carries out re-optimization to AUV task, and by optimum results reality When pass to each AUV, distribute follow-up task.
  2. 2. it is according to claim 1 a kind of for the more AUV task coordination methods of heterogeneous of execution sequence be present, its It is characterised by, described step (4) specifically includes：
    (4.1) mission mission area, AUV initial parameter information, initialization ant group algorithm basic parameter, time parameter are obtained； Mission mission area information includes the position coordinates in the marine site, oceanic area information；AUV parameter information mainly includes：AUV's Speed V2 when cruising speed V1, initial position message, execution clearance task；The basic parameter of ant group algorithm includes：Ant number Measure Num_ant, iterations iter_max, heuristic factor importance factor α, pheromones importance factor β, pheromones volatilization system Number ρ, Pheromone Matrix τ；
    (4.2) starting point vector sum terminal vector is set, and the numbering that starting point vector is each AUV, terminal vector is each task mesh Mark the numbering in marine site；
    (4.3) the state transition probability P of all possible paths is calculated according to starting point vector sum terminal vector；
    (4.4) paths are selected from state transition probability P using roulette method, the path as this circulation；
    (4.5) more ground zero vector sum emphasis vector, i.e., original starting point are substituted by terminal, and terminal originally is eliminated；
    (4.6) (4.3)-(4.5) are repeated, until terminal vector is sky；
    (4.7) (4.2)-(4.5) step is repeated, realizes and travels through until this generation all ants, record the road that each ant is passed by Footpath；
    (4.8) the pheromones τ on per paths is updated；
    (4.9) (4.2)-(4.8) are repeated until all generation iteration finish, the task distribution for recording the shortest time in all paths is believed Breath；
    The calculation formula of state transition probability P in step described above is as follows：
    <mrow> <msubsup> <mi>P</mi> <mrow> <mi>i</mi> <mi>j</mi> </mrow> <mi>k</mi> </msubsup> <mrow> <mo>(</mo> <mi>t</mi> <mo>,</mo> <mi>m</mi> <mo>)</mo> </mrow> <mo>=</mo> <mfenced open = "{" close = ""> <mtable> <mtr> <mtd> <mrow> <mfrac> <mrow> <msubsup> <mi>&amp;tau;</mi> <mrow> <mi>i</mi> <mi>j</mi> </mrow> <mi>&amp;alpha;</mi> </msubsup> <mrow> <mo>(</mo> <mi>t</mi> <mo>)</mo> </mrow> <msubsup> <mi>&amp;eta;</mi> <mrow> <mi>i</mi> <mi>j</mi> </mrow> <mi>&amp;beta;</mi> </msubsup> <mrow> <mo>(</mo> <mi>t</mi> <mo>,</mo> <mi>m</mi> <mo>)</mo> </mrow> </mrow> <mrow> <munder> <mo>&amp;Sigma;</mo> <mrow> <mi>s</mi> <mo>&amp;Element;</mo> <msub> <mi>allowed</mi> <mi>k</mi> </msub> </mrow> </munder> <msubsup> <mi>&amp;tau;</mi> <mrow> <mi>i</mi> <mi>s</mi> </mrow> <mi>&amp;alpha;</mi> </msubsup> <mrow> <mo>(</mo> <mi>t</mi> <mo>)</mo> </mrow> <msubsup> <mi>&amp;eta;</mi> <mrow> <mi>i</mi> <mi>s</mi> </mrow> <mi>&amp;beta;</mi> </msubsup> <mrow> <mo>(</mo> <mi>t</mi> <mo>)</mo> </mrow> </mrow> </mfrac> <mo>,</mo> </mrow> </mtd> <mtd> <mrow> <mi>j</mi> <mo>&amp;Element;</mo> <msub> <mi>allowed</mi> <mi>k</mi> </msub> </mrow> </mtd> </mtr> <mtr> <mtd> <mrow> <mn>0</mn> <mo>,</mo> </mrow> </mtd> <mtd> <mrow> <mi>o</mi> <mi>t</mi> <mi>h</mi> <mi>e</mi> <mi>r</mi> <mi>w</mi> <mi>i</mi> <mi>s</mi> <mi>e</mi> </mrow> </mtd> </mtr> </mtable> </mfenced> <mo>;</mo> </mrow>
    Heuristic function matrix η is the inverse of the time for the navigation that ant will be on the path in above formula, should be equation below：
    <mrow> <msub> <mi>&amp;eta;</mi> <mrow> <mi>i</mi> <mi>j</mi> </mrow> </msub> <mrow> <mo>(</mo> <mi>m</mi> <mo>)</mo> </mrow> <mo>=</mo> <mfrac> <mrow> <msub> <mi>V</mi> <mn>1</mn> </msub> <mrow> <mo>(</mo> <mi>m</mi> <mo>)</mo> </mrow> </mrow> <mrow> <mi>D</mi> <mrow> <mo>(</mo> <mi>i</mi> <mo>,</mo> <mi>j</mi> <mo>)</mo> </mrow> </mrow> </mfrac> <mo>;</mo> </mrow>
    Pheromone update principle formula in step described above is as follows：
    τij(t+1)=Δ τij(t,t+1)+ρ·τij(t)；
    <mrow> <msub> <mi>&amp;Delta;&amp;tau;</mi> <mrow> <mi>i</mi> <mi>j</mi> </mrow> </msub> <mrow> <mo>(</mo> <mi>t</mi> <mo>,</mo> <mi>t</mi> <mo>+</mo> <mn>1</mn> <mo>)</mo> </mrow> <mo>=</mo> <munderover> <mo>&amp;Sigma;</mo> <mrow> <mi>k</mi> <mo>=</mo> <mn>1</mn> </mrow> <mrow> <mi>N</mi> <mi>u</mi> <mi>m</mi> <mo>_</mo> <mi>a</mi> <mi>n</mi> <mi>t</mi> </mrow> </munderover> <msubsup> <mi>&amp;Delta;&amp;tau;</mi> <mrow> <mi>i</mi> <mi>j</mi> </mrow> <mi>k</mi> </msubsup> <mrow> <mo>(</mo> <mi>t</mi> <mo>,</mo> <mi>t</mi> <mo>+</mo> <mn>1</mn> <mo>)</mo> </mrow> <mo>;</mo> </mrow>
    I in above-mentioned each formula, j represent the starting and terminal point numbering in path respectively, and k numbers for ant, and m numbers for AUV, Num_ant For ant number, allowedkThe destination set for meeting constraints is represented, α is heuristic factor importance factor, and β is pheromones Importance factor, η are heuristic function matrix, and ρ is pheromones volatility coefficient, and its span is that 0~1, τ is Pheromone Matrix, D For the distance matrix of path Origin And Destination, V1For the velocity vector for the thunder type AUV that goes out, TijFor ant k hours underway.
  3. 3. it is according to claim 1 a kind of for the more AUV task coordination methods of heterogeneous of execution sequence be present, its It is characterised by, described step (6) specifically includes：
    (6.1) determine whether that new task object occurs, if do not occurred, keep original AUV task coordinate strategies；If Occur, obtain the coordinate information and quantity information of new task object, carry out following each step；
    (6.2) obtain each AUV working condition and it is expected that complete the time of task completed；Calculate and current time Time difference Delta_Time；Its Delta_Time=0 if AUV is not carried out task；
    (6.3) relevant information for the task that statistics is still not carried out, with emerging task object information together as this The task data of real-time optimization；
    (6.4) relevant parameter of ant group algorithm is initialized, starting point vector sum terminal vector is set, starting point vector is each AUV's Numbering, terminal vector are the numbering of each task object；
    (6.5) the state transition probability P of all possible paths is calculated according to starting point vector sum terminal vector；
    (6.6) paths are selected from state transition probability P using roulette method, the path as this circulation；
    (6.7) more ground zero vector sum terminal vector, i.e., original starting point are substituted by terminal, and terminal originally is eliminated；
    (6.8) (6.5)-(6.7) are repeated, until terminal vector is sky；
    (6.9) (6.4)-(6.7) are repeated, realizes and travels through until this generation all ants, record the path that each ant is passed by；
    (6.10) the pheromones τ on per paths is updated；
    (6.11) (6.4)-(6.10) are repeated until all generation iteration finish, the recording the shortest time in all paths of the task is distributed Information；
    Above-mentioned state transition probability P calculation formula is as follows：
    <mrow> <msubsup> <mi>P</mi> <mrow> <mi>i</mi> <mi>j</mi> </mrow> <mi>k</mi> </msubsup> <mrow> <mo>(</mo> <mi>t</mi> <mo>,</mo> <mi>m</mi> <mo>)</mo> </mrow> <mo>=</mo> <mfenced open = "{" close = ""> <mtable> <mtr> <mtd> <mrow> <mfrac> <mrow> <msubsup> <mi>&amp;tau;</mi> <mrow> <mi>i</mi> <mi>j</mi> </mrow> <mi>&amp;alpha;</mi> </msubsup> <mrow> <mo>(</mo> <mi>t</mi> <mo>)</mo> </mrow> <msubsup> <mi>&amp;eta;</mi> <mrow> <mi>i</mi> <mi>j</mi> </mrow> <mi>&amp;beta;</mi> </msubsup> <mrow> <mo>(</mo> <mi>t</mi> <mo>,</mo> <mi>m</mi> <mo>)</mo> </mrow> </mrow> <mrow> <munder> <mo>&amp;Sigma;</mo> <mrow> <mi>s</mi> <mo>&amp;Element;</mo> <msub> <mi>allowed</mi> <mi>k</mi> </msub> </mrow> </munder> <msubsup> <mi>&amp;tau;</mi> <mrow> <mi>i</mi> <mi>s</mi> </mrow> <mi>&amp;alpha;</mi> </msubsup> <mrow> <mo>(</mo> <mi>t</mi> <mo>)</mo> </mrow> <msubsup> <mi>&amp;eta;</mi> <mrow> <mi>i</mi> <mi>s</mi> </mrow> <mi>&amp;beta;</mi> </msubsup> <mrow> <mo>(</mo> <mi>t</mi> <mo>)</mo> </mrow> </mrow> </mfrac> <mo>,</mo> </mrow> </mtd> <mtd> <mrow> <mi>j</mi> <mo>&amp;Element;</mo> <msub> <mi>allowed</mi> <mi>k</mi> </msub> </mrow> </mtd> </mtr> <mtr> <mtd> <mrow> <mn>0</mn> <mo>,</mo> </mrow> </mtd> <mtd> <mrow> <mi>o</mi> <mi>t</mi> <mi>h</mi> <mi>e</mi> <mi>r</mi> <mi>w</mi> <mi>i</mi> <mi>s</mi> <mi>e</mi> </mrow> </mtd> </mtr> </mtable> </mfenced> <mo>;</mo> </mrow>
    η is the inverse of the time for the navigation that ant will be on the path in formula, and calculation formula is as follows：
    <mrow> <mfrac> <mn>1</mn> <mrow> <msub> <mi>&amp;eta;</mi> <mrow> <mi>i</mi> <mi>j</mi> </mrow> </msub> <mrow> <mo>(</mo> <mi>m</mi> <mo>)</mo> </mrow> </mrow> </mfrac> <mo>=</mo> <msub> <mi>T</mi> <mrow> <mi>i</mi> <mi>j</mi> </mrow> </msub> <mrow> <mo>(</mo> <mi>m</mi> <mo>)</mo> </mrow> <mo>=</mo> <mfrac> <mrow> <mi>D</mi> <mrow> <mo>(</mo> <mi>i</mi> <mo>,</mo> <mi>j</mi> <mo>)</mo> </mrow> </mrow> <mrow> <msub> <mi>V</mi> <mn>2</mn> </msub> <mrow> <mo>(</mo> <mi>m</mi> <mo>)</mo> </mrow> </mrow> </mfrac> <mo>+</mo> <mi>D</mi> <mi>e</mi> <mi>l</mi> <mi>t</mi> <mi>a</mi> <mo>_</mo> <mi>T</mi> <mi>i</mi> <mi>m</mi> <mi>e</mi> <mrow> <mo>(</mo> <mi>m</mi> <mo>)</mo> </mrow> <mo>;</mo> </mrow>
    I in above-mentioned formula, j represent the starting and terminal point numbering in path respectively, and k numbers for ant, and m numbers for AUV, allowedkTable Show the destination set for meeting constraints, α is heuristic factor importance factor, and β is pheromones importance factor, and η is inspiration letter Matrix number, ρ are pheromones volatility coefficient, and τ is Pheromone Matrix, and D is the distance matrix of path Origin And Destination, V2For the thunder type that goes out AUV velocity vector, Delta_Time are time difference vector, TijFor ant k hours underway.
  4. 4. it is according to claim 1 a kind of for the more AUV task coordination methods of heterogeneous of execution sequence be present, its It is characterised by：
    More AUV points of described heterogeneous are different types, and the AUV of each type configures different detection sensing equipments, each Sensing equipment can be used for completing different tasks, and AUV can also carry a variety of sensing equipments simultaneously；Described execution sequence problem tool Body is that mission area is different marine site, can be 1 independent task, can be also made up of N number of task, and it is certain to perform n-th task Using the implementing result of N-1 tasks as condition (N >=2), i.e., when the first task is after the completion of the execution in the marine site, could start The execution of second of task.
  5. 5. it is according to claim 1 a kind of for the more AUV task coordination methods of heterogeneous of execution sequence be present, its It is characterised by, described surface station specifically includes：
    Surface station has the task optimization distribution function of more AUV systems on working mother boat, and can pass through allocation result Radio or wireless network are sent；The heterogeneous AUV of configuration different sensors is mainly responsible for going to each marine site execution to appoint Business, each AUV can go up floatation surface and carry out data exchange with surface station, and data transfer is not needed between AUV.
  6. 6. it is according to claim 1 a kind of for the more AUV task coordination methods of heterogeneous of execution sequence be present, its It is characterised by, described object function specifically includes：
    Object function is as follows：
    MinT (m)=Tlength(m)+Tmission(m)+Twait(m)；
    M is AUV numbering in formula, and T (m) is m-th AUV general assignment duration, Tlength(m) it is m-th of AUV hours underway, Tmission(m) it is m-th of AUV task execution time, Twait(m) it is that part AUV is likely to occur before task terminates The time of wait state.
  7. 7. it is according to claim 1 a kind of for the more AUV task coordination methods of heterogeneous of execution sequence be present, its It is characterised by, described task coordinate model specifically includes：
    Described task coordinate model meets the time minimum under following each constraints；
    Constraints 1 is：Each single item task in task sequence must assure that execution and be only performed once；
    X in formulai,mFor task allocation information, i is mission number, and m numbers for AUV, and N is AUV quantity, and constraints 1 meets
    Constraints 2 is：That AUV is not used by but completing task can not be present；
    Used in formulamFor AUV service condition, constraints 2 is represented by usedm≥xi,m。
CN201710837778.1A 2017-09-18 2017-09-18 A kind of isomery more AUV system tasks coordination approach under task order constraint Pending CN107622348A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201710837778.1A CN107622348A (en) 2017-09-18 2017-09-18 A kind of isomery more AUV system tasks coordination approach under task order constraint

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201710837778.1A CN107622348A (en) 2017-09-18 2017-09-18 A kind of isomery more AUV system tasks coordination approach under task order constraint

Publications (1)

Publication Number Publication Date
CN107622348A true CN107622348A (en) 2018-01-23

Family

ID=61089791

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201710837778.1A Pending CN107622348A (en) 2017-09-18 2017-09-18 A kind of isomery more AUV system tasks coordination approach under task order constraint

Country Status (1)

Country Link
CN (1) CN107622348A (en)

Cited By (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110379247A (en) * 2019-07-19 2019-10-25 武汉理工大学 A kind of multitask, polygonal color ROV coorinated training analogue system and method
CN110456633A (en) * 2019-06-29 2019-11-15 西南电子技术研究所（中国电子科技集团公司第十研究所） Airborne multi-platform distributed task scheduling distribution method
CN111639811A (en) * 2020-06-01 2020-09-08 中国农业大学 Multi-agricultural-machine cooperative work remote management scheduling method based on improved ant colony algorithm
CN111898285A (en) * 2020-08-31 2020-11-06 西北工业大学 Underwater unmanned autonomous vehicle cluster operation efficiency evaluation method
CN112327838A (en) * 2020-10-29 2021-02-05 哈尔滨工程大学 Multi-unmanned surface vessel multi-task allocation method based on improved self-mapping algorithm
CN112859887A (en) * 2019-11-28 2021-05-28 中国科学院沈阳自动化研究所 Multi-underwater robot autonomous task allocation method based on space-based center
CN112862251A (en) * 2021-01-12 2021-05-28 西北工业大学 Distributed multi-AUV task allocation method based on improved PI algorithm
CN113033970A (en) * 2021-03-05 2021-06-25 西北工业大学 AUV cluster layered distribution method for large-scale tasks
CN114169804A (en) * 2022-01-07 2022-03-11 昆明理工大学 Optimized scheduling method for multi-distribution-center transport vehicle
CN114169707A (en) * 2021-11-24 2022-03-11 中国科学院沈阳自动化研究所 Task autonomous dynamic allocation method for multi-heterogeneous autonomous underwater robot

Cited By (20)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110456633A (en) * 2019-06-29 2019-11-15 西南电子技术研究所（中国电子科技集团公司第十研究所） Airborne multi-platform distributed task scheduling distribution method
CN110456633B (en) * 2019-06-29 2022-06-14 西南电子技术研究所（中国电子科技集团公司第十研究所） Airborne multi-platform distributed task allocation method
CN110379247B (en) * 2019-07-19 2021-12-07 武汉理工大学 Multitask and multi-role ROV (remote operated vehicle) collaborative training simulation system and method
CN110379247A (en) * 2019-07-19 2019-10-25 武汉理工大学 A kind of multitask, polygonal color ROV coorinated training analogue system and method
CN112859887B (en) * 2019-11-28 2022-06-14 中国科学院沈阳自动化研究所 Multi-underwater robot autonomous task allocation method based on space-based center
CN112859887A (en) * 2019-11-28 2021-05-28 中国科学院沈阳自动化研究所 Multi-underwater robot autonomous task allocation method based on space-based center
CN111639811A (en) * 2020-06-01 2020-09-08 中国农业大学 Multi-agricultural-machine cooperative work remote management scheduling method based on improved ant colony algorithm
CN111639811B (en) * 2020-06-01 2023-10-31 中国农业大学 Multi-agricultural-machine collaborative operation remote management scheduling method based on improved ant colony algorithm
CN111898285A (en) * 2020-08-31 2020-11-06 西北工业大学 Underwater unmanned autonomous vehicle cluster operation efficiency evaluation method
CN111898285B (en) * 2020-08-31 2022-07-26 西北工业大学 Underwater unmanned autonomous vehicle cluster operation efficiency evaluation method
CN112327838A (en) * 2020-10-29 2021-02-05 哈尔滨工程大学 Multi-unmanned surface vessel multi-task allocation method based on improved self-mapping algorithm
CN112327838B (en) * 2020-10-29 2023-03-31 哈尔滨工程大学 Multi-unmanned surface vessel multi-task allocation method based on improved self-mapping algorithm
CN112862251B (en) * 2021-01-12 2024-02-09 西北工业大学 Distributed multi-AUV task allocation method based on improved PI algorithm
CN112862251A (en) * 2021-01-12 2021-05-28 西北工业大学 Distributed multi-AUV task allocation method based on improved PI algorithm
CN113033970B (en) * 2021-03-05 2023-04-18 西北工业大学 AUV cluster layered distribution method for large-scale tasks
CN113033970A (en) * 2021-03-05 2021-06-25 西北工业大学 AUV cluster layered distribution method for large-scale tasks
CN114169707A (en) * 2021-11-24 2022-03-11 中国科学院沈阳自动化研究所 Task autonomous dynamic allocation method for multi-heterogeneous autonomous underwater robot
CN114169707B (en) * 2021-11-24 2024-08-09 中国科学院沈阳自动化研究所 Autonomous dynamic task allocation method for multi-heterogeneous autonomous underwater robot
CN114169804A (en) * 2022-01-07 2022-03-11 昆明理工大学 Optimized scheduling method for multi-distribution-center transport vehicle
CN114169804B (en) * 2022-01-07 2024-05-14 昆明理工大学 Optimized scheduling method for transport vehicles of multiple distribution centers

Similar Documents

Publication Publication Date Title
CN107622348A (en) A kind of isomery more AUV system tasks coordination approach under task order constraint
Gupta et al. Half a dozen real-world applications of evolutionary multitasking, and more
Abbasi et al. A cooperative dynamic task assignment framework for COTSBot AUVs
Yuan et al. A cooperative approach for multi-robot area exploration
Xin et al. A time-space network model for collision-free routing of planar motions in a multirobot station
Rupprecht et al. A survey for deep reinforcement learning in markovian cyber–physical systems: Common problems and solutions
Gul et al. Coordinated multi-robot exploration: Hybrid stochastic optimization approach
Petrovic et al. Can UAV and UGV be best buddies? Towards heterogeneous aerial-ground cooperative robot system for complex aerial manipulation tasks
Dewangan et al. A solution for priority-based multi-robot path planning problem with obstacles using ant lion optimization
CN112857370A (en) Robot map-free navigation method based on time sequence information modeling
Thompson et al. Robust mission planning for autonomous marine vehicle fleets
Chen et al. 2D multi-area coverage path planning using L-SHADE in simulated ocean survey
Huang et al. The USV path planning based on an improved DQN algorithm
D’urso et al. Hierarchical MCTS for scalable multi-vessel multi-float systems
BAYGIN et al. PSO based path planning approach for multi service robots in dynamic environments
Tang et al. Autonomous indoor mobile robot exploration based on wavefront algorithm
Xin et al. Simultaneous scheduling of processing machines and automated guided vehicles via a multi-view modeling-based hybrid algorithm
Han et al. Hybrid Algorithm-Based Full Coverage Search Approach With Multiple AUVs to Unknown Environments in Internet of Underwater Things
Gul et al. Efficient environment exploration for multi agents: A novel framework
Xu et al. A cell potential and motion pattern driven multi-robot coverage path planning algorithm
Miloradovic et al. Extended Colored Traveling Salesperson for Modeling Multi-Agent Mission Planning Problems.
CN112748729A (en) AGV route optimization and real-time scheduling method
Zhuo et al. A path planning strategy for data acquisition task using multiple autonomous underwater vehicles
Wang et al. Welding robot path optimization based on hybrid discrete PSO
Sandanika et al. Ros-based multi-robot system for efficient indoor exploration using a combined path planning technique

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
RJ01 Rejection of invention patent application after publication
RJ01 Rejection of invention patent application after publication

Application publication date: 20180123