CN102033536A - Scheduling, organization and cooperation system and method for multi-robot system - Google Patents

Scheduling, organization and cooperation system and method for multi-robot system Download PDF

Info

Publication number
CN102033536A
CN102033536A CN2010105997263A CN201010599726A CN102033536A CN 102033536 A CN102033536 A CN 102033536A CN 2010105997263 A CN2010105997263 A CN 2010105997263A CN 201010599726 A CN201010599726 A CN 201010599726A CN 102033536 A CN102033536 A CN 102033536A
Authority
CN
China
Prior art keywords
task
robot
management
evaluation result
resource
Prior art date
Application number
CN2010105997263A
Other languages
Chinese (zh)
Other versions
CN102033536B (en
Inventor
王占杰
Original Assignee
大连理工大学
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 大连理工大学 filed Critical 大连理工大学
Priority to CN2010105997263A priority Critical patent/CN102033536B/en
Publication of CN102033536A publication Critical patent/CN102033536A/en
Application granted granted Critical
Publication of CN102033536B publication Critical patent/CN102033536B/en

Links

Classifications

    • YGENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02PCLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
    • Y02P90/00Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
    • Y02P90/02Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]

Abstract

The invention discloses a scheduling, organization and cooperation system and a scheduling, organization and cooperation method for a multi-robot system. By adoption of distributive mode architecture, the information interaction is reduced, the systematic resources are used fully, the activity and the problem solving level of the overall system are enhanced, and the ability of intelligent coordination and flexible cooperation among a plurality of robots is enhanced. The method and the system intelligently perform scheduling and cooperation, reduce the communication traffic, reduce the systematic consumption, well adapt to a dynamic environment and complicated dynamic tasks, and enhance the problem solving capability by employing a dynamic decision maker scheme, namely every robot may be a task decision maker. According to the invention, a bulletin is used, a method of making a robot actively check the bulletin to acquire task information in an active free (capability) state is adopted and the selection of the best can be finished during checking, so that the information interaction is reduced, the information processing and comparison process is simplified, the communication traffic is greatly reduced and problems of information blocking and wastes of the systematic resources are solved.

Description

Cooperative system and method are organized in a kind of scheduling of multi-robot system

Technical field

Design robot control of the present invention field, particularly cooperative system and method are organized in a kind of scheduling of multi-robot system.

Background technology

Along with the continuous expansion in robot function and field, the flexible intelligent robot has replaced the robot of rigid, the function singleness of tradition, becomes current main direction of studying.Yet, in the multi-robot system, be in dynamically, between each robot in the uncertain environment, in the task implementation, need a large amount of information interactions and negotiation, yet along with the increase of robot quantity, communication pressure between robot and coordination difficulty also will increase greatly.And traditional multi-robot system communication exists a large amount of problems such as congestion information; Centralized or the layer-stepping pattern that is adopted on the organizational form in the dynamic change that adapts to main body ability and environment, has reduced the activity of total system, and bottleneck has been appearred in gerentocratic high request especially.In addition, also need to carry out the re-treatment and the comparison of bulk information when selecting the task executor, not only efficient is not high, but also has wasted a large amount of system resource.

The research of multi-robot system at present has been subjected to common attention, and relatively more classical experimental system has: CEBOT (Cellular Robotic system), ALLIANCE/L-ALLANCE system, SWARM system.The colony's architecture, perception that these researchs relate to multirobot and Multi-sensor Fusion, communicate by letter and many aspects such as the coordination system, Task Distribution, study, the realization of motion planning system, and are main contents of multi-robot system research for the research of the coordination and cooperation of a plurality of robots.The scholar that external current research mainly contains U.S. USC university sets up soeiallyM0bile and TheNerdHerd experimental system; The Collective Robotics(collective robot system of people such as the C.R.Kube development of Canada Alberta university) and many work of doing at aspects such as multi-Agent cooperation, autonomous Agent architecture, multi-Agent communications of doctor LynneE.Parker of U.S. Oak Ridge National Laboratory.With respect to abroad, China starts late for the research of multi-robot system, generally concentrate on aspects such as multirobot cooperation and multirobot study, and spininess is to the application foundation Journal of Sex Research of intelligent manufacturing system and robot soccer.The MACAS system that main contents have Chinese Academy of Sciences's Shenyang Institute of Automation to set up; The multiple-mobile-robot system platform ActivMedia Robotics of Shanghai Communications University's development.Yet all multi-methods that these researchs are adopted during the problem aspect Task Distribution and the cooperation in handling the multirobot dispatching system are in the dynamic change that adapts to main body ability and environment, reduce the traffic, the problem solving ability that improves total system and the aspects such as overall performance optimization of activity and still exist problems.The design and in conjunction with new application background and technology, has designed new the flexible tissue model and the smart collaboration scheme that are applicable to the multirobot scheduling on the basis of accumulation available research achievements, this scheme will have broad application prospects.

Summary of the invention

The objective of the invention is: at all bottlenecks that exist in the current multi-robot system, designed a kind of new dispatch group organization model that is suitable for the flexible robot, and a kind of smart collaboration mechanism has been proposed based on model, to reduce the traffic, minimizing system overhead, the raising problem solving ability in the scheduling process, improve the entire system performance.

The technical solution used in the present invention is: provide a kind of scheduling of multi-robot system to organize collaboration method, may further comprise the steps:

The first step, task issuing process: after the management and running unit of first Robot1 of robot finds that new task arrives, task management unit, rm-cell and management and running unit associations are made self-assessment, provide evaluation result, first Robot1 of robot becomes the decision maker, and the management and running unit together is published to announcement board Bulletin with mission bit stream together with evaluation result.

Second step, executor's selection course: selecting the task executor before the time arrives, there is the robot of idling-resource to check the mission bit stream of announcement board Bulletin, carry out self-assessment, the evaluation result comparison module of management and running unit compares evaluation result and current decision maker's evaluation result, if this time evaluation result is better than current decision maker's evaluation result, the evaluation result of then revising in the current task information is new evaluation result, and replacing current decision maker becomes new decision maker, and notifies former decision maker to abandon right to make decision to this task; Otherwise, abandon this task.

The 3rd step, executor's deterministic process: if task executor's selection course finishes, and this moment, the decision maker still was discoverer first Robot1 of robot of task, the task ability judge module of management and running unit judges that first Robot1 of robot possesses the ability of finishing this task, then current decision maker becomes the task executor, begin task executions, and upgrade self-ability information with regard to the performance of task.

Wherein, described the 3rd step further comprises: if task executor's selection course finishes, and this moment, the decision maker still was discoverer first Robot1 of robot of task, and first Robot1 of robot does not possess the ability of finishing this task, then this task is decomposed by the task decomposing module of the management and running unit of first Robot1 of robot, bear part task wherein, and the task that will remain is issued bulletin, three steps of the repetition first step to the once more.

Wherein, the step of described self-assessment comprises:

At first, after new task is found in the management and running unit, management and running unit and task management unit and resource management module are held consultation: if this task directly obtains from real world, then estimate according to the current task load information of task management unit calculating and the resource load information of rm-cell calculating, then evaluation result is issued together with mission bit stream by its management and running unit; So that the other machines people is known new information, this robot also becomes current decision maker, and this condition of information is monitored in beginning in real time.If receive the notice that the other machines people sends, then abandon right to make decision to this task.

Secondly, task management unit and resource management module carry out the calculating of pre-service and processing power to current task.

Wherein, preprocessing process is: the mission bit stream collection module collection monitoring in the task management unit and all mission bit streams of tracking comprise: priority Ui, beginning time limit Si, processing time Ti, carry out required resource Ri etc. with traffic summation Ci, the current place node Ni of other task, task; Resource information monitoring module dynamic collection in the resource management module is the monitoring resource information relevant with processing power in real time, comprising: the flow Ii of controller utilization factor Pi, instruction ready queue length L i, storage space Mi, each interface and utilization factor Ei etc.

Wherein, processing power computation process is: mission bit stream computing module in the task management unit is according to all mission bit streams of collecting, and with they integrate calculate current task loading condition T_load=μ 1Ui+μ 2Si+μ 3Ti+μ 4Ci+μ 5Ni+μ 6Ri+ Resource information computing module in the rm-cell is according to the information of dynamic real-time monitor and collection, and with they integrate calculate current resource load situation R_load=β 1Pi+β 2Li+β 3Mi+β 4Ii+β 5Ei+ Wherein, μ i, β i are respectively the wherein weights of each information, , , can specifically set according to different demands.

At last, the management and running unit is according to negotiation result, and in conjunction with other factors, comprising: antistress and communication capacity, the discovery of task and the ability of acceptance, and the issue capability of task; By finishing the overall evaluation work of ability, and provide final self evaluation result F=f(T_ load with the communication negotiation of other unit, R_load ...), evaluation result F is the function of each factor.

Other provides a kind of scheduling of multi-robot system to organize cooperative system, and described scheduling organizes cooperative system to comprise:

The task management unit, be used for monitoring and follow the tracks of all mission bit streams, described mission bit stream comprises priority, beginning time limit, processing time, the traffic summation with other task, current place node, the required resource of task execution, and mission bit stream integrated calculate current task loading condition, according to mission requirements, the application resource is carried out its task.

Rm-cell is used to manage all resources relevant with processing power, the monitoring of dynamic real-time and collect resource information, and resource information is integrated the current resource load situation of calculating; Described resource comprises: the flow of controller utilization factor, instruction ready queue length, storage space, each interface and utilization factor etc.

The management and running unit: be used to the discovery and the acceptance of the task that realizes, and the issue of task; Hold consultation with its task management unit, rm-cell, finish the overall evaluation work of ability, and provide evaluation result; Realize with the other machines people between communicate by letter; After receiving an assignment, can carry out task scheduling, finish the decomposition of extensive task according to task issue situation.

Described task management unit, rm-cell and management and running unit interconnect by communication interface.

Wherein, described management and running unit comprises:

The evaluation result comparison module, be used for self evaluation result and current decision maker's evaluation result is compared: if evaluation result is better than former decision maker, revise then that evaluation result is new evaluation result in the announcement board, replace former decision maker and become new decision maker, give notice to former decision maker, so that former decision maker abandons the right to make decision to this task, otherwise do not participate in the competition.

The task ability judge module, be used to judge whether first Robot1 of robot possesses the ability of finishing this task: arrived uncontested person when the effective time of message, the task ability judge module is made a strategic decision according to its ability, and ability is enough, then carry out its task, otherwise task is decomposed.

The release module of task is issued evaluation result together with mission bit stream, so that the other machines people is known new information.

Wherein, described management and running unit further comprises:

The task decomposing module, be used for extensive task is decomposed: arrived uncontested person when the effective time of message, first Robot1 of robot does not possess the ability of finishing this task again, then task is decomposed, carry out the part task after decomposing, and the task that will remain is handled by this scheduling coordination mechanism once more.

Wherein, described task management unit comprises:

The mission bit stream collection module, all mission bit streams that are used to monitor and follow the tracks of comprise: priority, beginning time limit, processing time, the traffic summation with other task, current place node, task are carried out required resource etc.

The mission bit stream computing module is used for according to all mission bit streams of collecting, with they integrate calculate current task loading condition T_load=μ 1Ui+μ 2Si+μ 3Ti+μ 4Ci+μ 5Ni+μ 6Ri+ Wherein, μ i is weights, , specifically set according to different needs.

Wherein, described rm-cell comprises:

The resource information monitoring module is used to manage all resources relevant with processing power, comprising: the flow of controller utilization factor, instruction ready queue length, storage space, each interface and utilization factor etc.

The resource information computing module is used for the information according to dynamic real-time monitor and collection, and with they integrate calculate current resource load situation R_load=β 1Pi+β 2Li+β 3Mi+β 4Ii+β 5Ei+ Wherein, β i is weights, , specifically set according to different needs.

Beneficial technical effects of the present invention:

(1) dispatch group organization model that the design proposed and coordination mechanism are a kind of brand-new design proposals, distributed pattern architecture is adopted in design, can reduce information interaction, make full use of system resource, improve the activity and the problem solving level of total system, strengthened the ability of the intelligent coordinated and flexible cooperation between a plurality of robots.

(2) along with the variation of task and environment, the role of each robot also can change, develop to flexible intelligent gradually from the robot of function singleness, cause the dynamic change of robot system institutional framework, to this design person's scheme that adopts the dynamic decision, be the decision maker that each robot all may become task, intelligent dispatching and cooperating, reach the reduction traffic, reduce the purpose of system consumption, and can better adapt to dynamic environment and complicated dynamic task, improve the ability of finding the solution of problem.

(3) adopt the method for network service in the traditional scheme between each robot, can take a large amount of Internet resources, therefore, cause the obstruction of network when limited or network resource loads is heavier at Internet resources easily.We have quoted announcement board in this article at this defective, employing allows robot check initiatively that when the resource free time (ability) is arranged announcement board obtains the method for mission bit stream, simultaneously, the process of checking can be finished the selection for the superior, thereby minimizing information interaction, simplified the information processing comparison procedure in the former traditional scheme, greatly reduced the traffic, effectively solved the information that occurs in the traditional scheme and block and the resource waste problem.

Description of drawings

The dispatch group organization model figure of Fig. 1 multi-robot system of the present invention.

The cooperative system frame diagram is organized in the scheduling of Fig. 2 multi-robot system of the present invention.

The task management unit framework figure of cooperative system is organized in the scheduling of Fig. 3 multi-robot system of the present invention.

The rm-cell frame diagram of cooperative system is organized in the scheduling of Fig. 4 multi-robot system of the present invention.

The management and running unit framework figure of cooperative system is organized in the scheduling of Fig. 5 multi-robot system of the present invention.

Collaboration method task issuing process process flow diagram is organized in the scheduling of Fig. 6 multi-robot system of the present invention.

Collaboration method executor selection course process flow diagram is organized in the scheduling of Fig. 7 multi-robot system of the present invention.

Collaboration method executor deterministic process process flow diagram is organized in the scheduling of Fig. 8 multi-robot system of the present invention.

Embodiment

The present invention is described in detail below in conjunction with drawings and Examples.

One, thinking of the present invention:

Development along with computing technique, the continuous expansion of the deep and application of multirobot technical research, the quantity size of robot application is also in continuous expansion, the function that robot has is also enriched constantly, communication pressure between the robot and coordination difficulty will strengthen, make to become more and more important about the formation of robot tissue and the research of organizing based on robot of finding the solution problems such as mechanism, how intelligent.This mechanism proposes thoughts such as distributed group organization model, announcement board, dynamic decision person and will reduce the traffic in the multi-robot system for how to solve, improves problem solving ability, improves resource utilization, better adapt to difficult problems such as dynamic environment and complicated dynamic task and represented wide prospect.

The traffic is big, resource utilization is not high, shortcomings such as the problem ability is lower are found the solution by system because the scheduling of current multirobot exists, and therefore, the design organize models adopts distributed operating strategy, resource that can the better utilization system.In new intelligence flexible dispatch group organization model and coordination mechanism, we have introduced the means of announcement board as decision maker's release task message, take to allow the robot of resource free time (ability) check that initiatively announcement board is to obtain the mode of mission bit stream, to make full use of system resource, to reduce the traffic.And adopt dynamic decision person strategy, reduce information interaction, simplify information process, better adapt to dynamic environment and complicated dynamic task.

Two, organize models:

The design's organize models wherein all comprises three parts, task management, resource management and behavior management in the dispatch group organization model of each robot and the coordination mechanism as shown in Figures 1 and 2.Each several part all can adopt prior art to be realized, it is that example describes that the design adopts the Agent technology, and the each several part function is described below respectively:

Task management Agent(TA), be task management unit 100: monitor and follow the tracks of all mission bit streams, these information comprise that priority, beginning time limit, processing time, the traffic summation with other task, current place node, task carry out required resource etc., and they can be integrated and calculate current task loading condition.Can reasonably apply for resource according to mission requirements, carry out its task.As shown in Figure 3, task management Agent(TA) comprising: mission bit stream collection module 110, all mission bit streams that are used to monitor and follow the tracks of comprise: priority, beginning time limit, processing time, the traffic summation with other task, current place node, task are carried out required resource etc.; And mission bit stream computing module 120, be used for according to all mission bit streams of collecting, with they integrate calculate current task loading condition T_load=μ 1Ui+μ 2Si+μ 3Ti+μ 4Ci+μ 5Ni+μ 6Ri+ Wherein, μ i is weights, , specifically set according to different needs.

Resource management Agent(RA), be rm-cell 200: manage all resources relevant with processing power, generally include: the flow of controller utilization factor, instruction ready queue length, storage space, each interface and utilization factor etc., the monitoring of dynamic real-time and collect these information, and they are integrated the current resource load situation of calculating.As shown in Figure 4, resource management Agent(RA) comprising: resource information monitoring module 210, be used to manage all resources relevant, comprise: the flow of controller utilization factor, instruction ready queue length, storage space, each interface and utilization factor etc. with processing power; Resource information computing module 220 is used for the information according to dynamic real-time monitor and collection, and with they integrate calculate current resource load situation R_load=β 1Pi+β 2Li+β 3Mi+β 4Ii+β 5Ei+ Wherein, β i is weights, , specifically set according to different needs.

Management and running Agent(SA), promptly the management and running unit 300: should have rapid reaction and communication capacity, realize the discovery and the acceptance of task, and the issue of task; Can hold consultation with other Agent, finish the overall evaluation work of ability, and provide evaluation result; Realize with the other machines people between work such as communicate by letter.Particularly after receiving an assignment, can carry out task scheduling, also can finish the disintegration of extensive task according to task issue situation.As shown in Figure 5, management and running Agent(SA) comprising: evaluation result comparison module 310, be used for self evaluation result and current decision maker's evaluation result is compared: if evaluation result is better than former decision maker, revise then that evaluation result is new evaluation result in the announcement board, replace former decision maker and become new decision maker, give notice to former decision maker,, otherwise do not participate in the competition so that former decision maker abandons the right to make decision to this task; Task ability judge module 320, be used to judge whether first Robot1 of robot possesses the ability of finishing this task: arrived uncontested person when the effective time of message, task ability judge module 320 is made a strategic decision according to its ability, and ability is enough, then carry out its task, otherwise task is decomposed; Task decomposing module 330, be used for extensive task is decomposed: arrived uncontested person when the effective time of message, first Robot1 of robot does not possess the ability of finishing this task again, then task is decomposed, carry out the part task after decomposing, and the task that will remain is handled by this scheduling coordination mechanism once more; And task release module 340, evaluation result is issued together with mission bit stream, so that the other machines people is known new information.

Three, coordination mechanism:

The new task discovery mechanism: a kind of is the new task that obtains issue, and another kind is the arrival of new task in the real world, and this mechanism is mainly born by management and running Agent.After finding new task, management and running Agent holds consultation immediately and between other Agent, finish for the pre-service of this mission bit stream and the calculating of processing power etc. by resource management Agent and task management Agent, management and running Agent then comes its ability of comprehensive evaluation according to negotiation result and in conjunction with other factors, and provides evaluation result.This process is also referred to as self and estimates, and enters the new task processing procedure afterwards.

New task processing procedure: if obtaining of task is the new task of issue, then management and running Agent need compare evaluation result and task publisher's evaluation result, if evaluation result is better than former decision maker, revise then that evaluation result is new evaluation result in the announcement board, replace former decision maker and become new decision maker, give notice to former decision maker,, otherwise do not participate in the competition so that former decision maker abandons the right to make decision to this task.

If this task directly obtains from real world, then evaluation result is issued together with mission bit stream by its management and running Agent, so that the other machines people is known new information, this robot also becomes current decision maker, and this condition of information is monitored in beginning in real time.If receive the notice that the other machines people sends, then abandon right to make decision to this task.

Decision-making mechanism: arrived uncontested person the effective time as message, and management and running Agent makes a strategic decision according to its ability, and ability is enough carried out its task; Otherwise task is decomposed, carry out the part task after decomposing, and the task that will remain is handled by this scheduling coordination mechanism once more.

Four, this model workflow simply is described below:

(1) after Robot1 finds that new task arrives, at first make self-assessment, provide evaluation result, mission bit stream is together released together with evaluation result, as shown in Figure 6, (a) task issuing process.

(2) selecting the task executor before the time arrives (being in the effective time of message), there is the robot of idling-resource (ability) to check mission bit stream, carry out self-assessment, and evaluation result and current decision maker's evaluation result compared, if this evaluation result is better than current decision maker, the evaluation result of then revising in the current task information is new evaluation result, and replace current decision maker and become new decision maker, and notify former decision maker to abandon right to make decision to this task, as shown in Figure 7, (b) executor's selection course.

(3) if task executor's selection course finishes, and this moment, the decision maker still was the discoverer Robot1 of task, and Robot1 do not possess the ability of finishing this task, then by Robot1 this task decomposed, bear part task wherein, and the task that will remain issues bulletin once more, otherwise then current decision maker becomes the task executor, the beginning task executions, and upgrade self-ability information with regard to the performance of task, and as shown in Figure 8, (C) executor's deterministic process.

Above content be in conjunction with optimal technical scheme to further describing that the present invention did, can not assert that the concrete enforcement of invention only limits to these explanations.Concerning the general technical staff of the technical field of the invention, under the prerequisite that does not break away from design of the present invention, can also make simple deduction and replacement, all should be considered as protection scope of the present invention.

Claims (8)

1. collaboration method is organized in the scheduling of a multi-robot system, may further comprise the steps:
Steps A, task issuing process: after the management and running unit (300) of first Robot1 of robot finds that new task arrives, task management unit (100), rm-cell (200) and management and running unit (300) are united and are made self-assessment, provide evaluation result, first Robot1 of robot becomes the decision maker, and management and running unit (300) together are published to announcement board Bulletin with mission bit stream together with evaluation result;
Step B, executor's selection course: selecting the task executor before the time arrives, there is the robot of idling-resource to check the mission bit stream of announcement board Bulletin, carry out self-assessment, the evaluation result comparison module (310) of management and running unit (300) compares evaluation result and current decision maker's evaluation result, if this time evaluation result is better than current decision maker's evaluation result, the evaluation result of then revising in the current task information is new evaluation result, and replacing current decision maker becomes new decision maker, and notifies former decision maker to abandon right to make decision to this task; Otherwise, abandon this task;
Step C, executor's deterministic process: if task executor's selection course finishes, and this moment, the decision maker still was discoverer first Robot1 of robot of task, the task ability judge module (320) of management and running unit (300) judges that first Robot1 of robot possesses the ability of finishing this task, then current decision maker becomes the task executor, begin task executions, and upgrade self-ability information with regard to the performance of task.
2. collaboration method is organized in the scheduling of multi-robot system according to claim 1, described step C further comprises: if task executor's selection course finishes, and this moment, the decision maker still was discoverer first Robot1 of robot of task, and first Robot1 of robot does not possess the ability of finishing this task, then this task is decomposed by the task decomposing module (330) of the management and running unit (300) of first Robot1 of robot, bear part task wherein, and the task that will remain issues bulletin once more, and repeating step A is to step C.
3. collaboration method is organized in the scheduling of multi-robot system according to claim 1, and the step of described self-assessment comprises:
At first, after management and running unit (300) find new task, hold consultation with task management unit (100) and resource management module (200) in management and running unit (300): if this task directly obtains from real world, then estimate according to the current task load information of task management unit (100) calculating and the resource load information of rm-cell (200) calculating, then evaluation result is issued together with mission bit stream by its management and running unit (300); So that the other machines people is known new information, this robot also becomes current decision maker, and this condition of information is monitored in beginning in real time;
If receive the notice that the other machines people sends, then abandon right to make decision to this task;
Secondly, task management unit (100) and resource management module (200) carry out the calculating of pre-service and processing power to current task;
Wherein, preprocessing process is: mission bit stream collection module (110) collection monitoring in task management unit (100) and all mission bit streams of tracking comprise: priority Ui, beginning time limit Si, processing time Ti, traffic summation Ci, current place node Ni with other task, the required resource Ri of task execution; Resource information monitoring module (210) dynamic collection in the resource management module (200) is the monitoring resource information relevant with processing power in real time, comprising: the flow Ii and the utilization factor Ei of controller utilization factor Pi, instruction ready queue length L i, storage space Mi, each interface;
Wherein, processing power computation process is: mission bit stream computing module (120) in task management unit (100) is according to all mission bit streams of collecting, and with they integrate calculate current task loading condition T_load=μ 1Ui+μ 2Si+μ 3Ti+μ 4Ci+μ 5Ni+μ 6Ri+ Resource information computing module (220) in the rm-cell (200) is according to the information of dynamic real-time monitor and collection, and with they integrate calculate current resource load situation R_load=β 1Pi+β 2Li+β 3Mi+β 4Ii+β 5Ei+ Wherein, μ i, β i are respectively the wherein weights of each information, , , occurrence can specifically be set according to different demands;
At last, management and running unit (300) according to negotiation result, and in conjunction with other factors, comprising: antistress and communication capacity, the discovery of task and the ability of acceptance, and the issue capability of task; By finishing the overall evaluation work of ability, and provide final self evaluation result F=f(T_ load with the communication negotiation of other unit, R_load ...), evaluation result F is T_ load, the function of factors such as R_load.
4. cooperative system is organized in the scheduling of a multi-robot system, it is characterized in that, described scheduling organizes cooperative system to comprise:
Task management unit (100), be used for monitoring and follow the tracks of all mission bit streams, described mission bit stream comprises priority, beginning time limit, processing time, the traffic summation with other task, current place node, the required resource of task execution, and mission bit stream integrated calculate current task loading condition, according to mission requirements, the application resource is carried out its task;
Rm-cell (200) is used to manage all resources relevant with processing power, the monitoring of dynamic real-time and collect resource information, and resource information is integrated the current resource load situation of calculating; Described resource comprises: the flow and the utilization factor of controller utilization factor, instruction ready queue length, storage space, each interface;
Management and running unit (300): be used to the discovery and the acceptance of the task that realizes, and the issue of task; Hold consultation with its task management unit (100), rm-cell (200), finish the overall evaluation work of ability, and provide evaluation result; Realize with the other machines people between communicate by letter; After receiving an assignment, can carry out task scheduling, finish the decomposition of extensive task according to task issue situation;
Described task management unit (100), rm-cell (200) and management and running unit (300) interconnect by communication interface.
5. cooperative system is organized in the scheduling of multi-robot system according to claim 4, it is characterized in that, described management and running unit (300) comprising:
Evaluation result comparison module (310), be used for self evaluation result and current decision maker's evaluation result is compared: if evaluation result is better than former decision maker, revise then that evaluation result is new evaluation result in the announcement board, replace former decision maker and become new decision maker, give notice to former decision maker, so that former decision maker abandons the right to make decision to this task, otherwise do not participate in the competition;
Task ability judge module (320), be used to judge whether first Robot1 of robot possesses the ability of finishing this task: arrived uncontested person when the effective time of message, task ability judge module (320) is made a strategic decision according to its ability, ability is enough, then carry out its task, otherwise task is decomposed;
The release module of task (340) is issued evaluation result together with mission bit stream, so that the other machines people is known new information.
6. cooperative system is organized in the scheduling of multi-robot system according to claim 5, it is characterized in that, described management and running unit (300) further comprises:
Task decomposing module (330), be used for extensive task is decomposed: arrived uncontested person when the effective time of message, first Robot1 of robot does not possess the ability of finishing this task again, then task is decomposed, carry out the part task after decomposing, and the task that will remain is handled by this scheduling coordination mechanism once more.
7. cooperative system is organized in the scheduling of multi-robot system according to claim 4, it is characterized in that, described task management unit (100) comprising:
Mission bit stream collection module (110), all mission bit streams that are used to monitor and follow the tracks of comprise: priority, beginning time limit, processing time, the traffic summation with other task, current place node, task are carried out required resource;
Mission bit stream computing module (120) is used for according to all mission bit streams of collecting, with they integrate calculate current task loading condition T_load=μ 1Ui+μ 2Si+μ 3Ti+μ 4Ci+μ 5Ni+μ 6Ri+ Wherein, μ i is weights, , specifically set according to different needs.
8. cooperative system is organized in the scheduling of multi-robot system according to claim 4, it is characterized in that, described rm-cell (200) comprising:
Resource information monitoring module (210) is used to manage all resources relevant with processing power, comprising: the flow and the utilization factor of controller utilization factor, instruction ready queue length, storage space, each interface;
Resource information computing module (220) is used for the information according to dynamic real-time monitor and collection, and with they integrate calculate current resource load situation R_load=β 1Pi+β 2Li+β 3Mi+β 4Ii+β 5Ei+ Wherein, β i is weights, , specifically set according to different needs.
CN2010105997263A 2010-12-22 2010-12-22 Scheduling, organization and cooperation system and method for multi-robot system CN102033536B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN2010105997263A CN102033536B (en) 2010-12-22 2010-12-22 Scheduling, organization and cooperation system and method for multi-robot system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN2010105997263A CN102033536B (en) 2010-12-22 2010-12-22 Scheduling, organization and cooperation system and method for multi-robot system

Publications (2)

Publication Number Publication Date
CN102033536A true CN102033536A (en) 2011-04-27
CN102033536B CN102033536B (en) 2012-12-26

Family

ID=43886548

Family Applications (1)

Application Number Title Priority Date Filing Date
CN2010105997263A CN102033536B (en) 2010-12-22 2010-12-22 Scheduling, organization and cooperation system and method for multi-robot system

Country Status (1)

Country Link
CN (1) CN102033536B (en)

Cited By (23)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102499728A (en) * 2011-12-02 2012-06-20 苏州生物医学工程技术研究所 Medical miniature multi-robot system with self-organizing ability
CN102662350A (en) * 2012-05-31 2012-09-12 东南大学 Track teaching and planning method of master-slave mode multi-robot cooperative system
CN103064743A (en) * 2012-12-27 2013-04-24 深圳先进技术研究院 Resource scheduling method for multiple robots and resource scheduling system thereof
CN103176454A (en) * 2013-02-21 2013-06-26 深圳先进技术研究院 Clustering method and clustering system of multi-robot system
CN103235595A (en) * 2013-04-27 2013-08-07 湖南科技大学 Control system and control method of outdoor micro ground swarm robot
CN103324173A (en) * 2013-06-08 2013-09-25 上海交通大学 Multiple-heterogeneous robot cooperation method based on middleware
CN103399543A (en) * 2013-07-23 2013-11-20 清华大学 Method and system for scheduling multiple robot equipment cooperated for transferring wafers
WO2014101558A1 (en) * 2012-12-27 2014-07-03 深圳先进技术研究院 Processing method and system for controlling service amounts of multiple robots
CN104820621A (en) * 2015-04-27 2015-08-05 南京大学 Intelligent vehicle cooperation method based on distributed shared memory
CN105069530A (en) * 2015-08-13 2015-11-18 肇庆学院 Multi-robot task allocation method based on multi-objective optimization
CN105093934A (en) * 2015-08-17 2015-11-25 哈尔滨工业大学 Distributed finite time tracking control method for multi-robot system in view of interference and model uncertainty
CN105182828A (en) * 2015-09-18 2015-12-23 深圳前海达闼科技有限公司 Method for equipment to request assistance, method for responding to equipment assistance request, and apparatuses
CN105488892A (en) * 2016-01-04 2016-04-13 杭州亚美利嘉科技有限公司 Method and server for robot queuing management
CN105553017A (en) * 2015-12-31 2016-05-04 小米科技有限责任公司 Robot charging system
CN105729491A (en) * 2016-04-01 2016-07-06 纳恩博(北京)科技有限公司 Executing method, device and system for robot task
CN106557362A (en) * 2016-11-15 2017-04-05 北京光年无限科技有限公司 It is applied to the task immigration method and system of intelligent robot
CN106777135A (en) * 2016-05-27 2017-05-31 中科鼎富(北京)科技发展有限公司 Service scheduling method, device and robot service system
CN107077651A (en) * 2016-12-29 2017-08-18 深圳前海达闼云端智能科技有限公司 Robot cooperated method, device, robot and computer program product
CN107223261A (en) * 2016-12-07 2017-09-29 深圳前海达闼云端智能科技有限公司 Man-machine complex decision method and apparatus
CN107247460A (en) * 2017-06-01 2017-10-13 三峡大学 A kind of cluster control method and system of machine honeybee
CN107450491A (en) * 2017-08-04 2017-12-08 杭州南江机器人股份有限公司 A kind of robot scheduling system, method, electronic equipment and storage medium
CN108171394A (en) * 2017-11-21 2018-06-15 中国科学院自动化研究所 Multi-robot Task Allocation based on hierachical structure and resource consolidation
WO2020103298A1 (en) * 2018-11-19 2020-05-28 炬星科技(深圳)有限公司 Distributed robot dispatch decision-making method, device and system, electronic device and storage medium

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP3588405A1 (en) * 2018-06-29 2020-01-01 Tata Consultancy Services Limited Systems and methods for scheduling a set of non-preemptive tasks in a multi-robot environment

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2004104716A1 (en) * 2003-05-16 2004-12-02 Fsi International, Inc. Scheduling multi-robot processing systems
CN101551661A (en) * 2009-05-12 2009-10-07 广东工业大学 Control method for multi-robot system

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2004104716A1 (en) * 2003-05-16 2004-12-02 Fsi International, Inc. Scheduling multi-robot processing systems
CN101551661A (en) * 2009-05-12 2009-10-07 广东工业大学 Control method for multi-robot system

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
《微计算机信息》 20051231 王占杰 等 "多agent动态调度控制模型" 第21卷, 第1期 *

Cited By (35)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102499728A (en) * 2011-12-02 2012-06-20 苏州生物医学工程技术研究所 Medical miniature multi-robot system with self-organizing ability
CN102662350A (en) * 2012-05-31 2012-09-12 东南大学 Track teaching and planning method of master-slave mode multi-robot cooperative system
CN102662350B (en) * 2012-05-31 2013-11-27 东南大学 Track teaching and planning method of master-slave mode multi-robot cooperative system
CN103064743A (en) * 2012-12-27 2013-04-24 深圳先进技术研究院 Resource scheduling method for multiple robots and resource scheduling system thereof
CN103064743B (en) * 2012-12-27 2017-09-26 深圳先进技术研究院 A kind of resource regulating method and its resource scheduling system for multirobot
WO2014101558A1 (en) * 2012-12-27 2014-07-03 深圳先进技术研究院 Processing method and system for controlling service amounts of multiple robots
CN103176454A (en) * 2013-02-21 2013-06-26 深圳先进技术研究院 Clustering method and clustering system of multi-robot system
CN103176454B (en) * 2013-02-21 2015-07-29 深圳先进技术研究院 The cluster method of multi-robot system and system
CN103235595A (en) * 2013-04-27 2013-08-07 湖南科技大学 Control system and control method of outdoor micro ground swarm robot
CN103235595B (en) * 2013-04-27 2016-03-16 湖南科技大学 A kind of outdoor micro ground swarm robot control system and control method
CN103324173A (en) * 2013-06-08 2013-09-25 上海交通大学 Multiple-heterogeneous robot cooperation method based on middleware
CN103324173B (en) * 2013-06-08 2015-10-28 上海交通大学 Based on many heterogeneous robots collaboration method of middleware
CN103399543B (en) * 2013-07-23 2016-03-23 清华大学 The dispatching method of multiple robot devices of film trasport cooperation and system
CN103399543A (en) * 2013-07-23 2013-11-20 清华大学 Method and system for scheduling multiple robot equipment cooperated for transferring wafers
CN104820621A (en) * 2015-04-27 2015-08-05 南京大学 Intelligent vehicle cooperation method based on distributed shared memory
CN104820621B (en) * 2015-04-27 2018-07-24 南京大学 Intelligent carriage Synergistic method based on Distributed shared memory
CN105069530A (en) * 2015-08-13 2015-11-18 肇庆学院 Multi-robot task allocation method based on multi-objective optimization
CN105069530B (en) * 2015-08-13 2019-02-26 肇庆学院 A kind of multi-robot Task Allocation based on multiple-objection optimization
CN105093934A (en) * 2015-08-17 2015-11-25 哈尔滨工业大学 Distributed finite time tracking control method for multi-robot system in view of interference and model uncertainty
CN105093934B (en) * 2015-08-17 2017-08-25 哈尔滨工业大学 Consider interference and the distributed finite time tracking controller design method of multi-robot system of model uncertainty
CN105182828A (en) * 2015-09-18 2015-12-23 深圳前海达闼科技有限公司 Method for equipment to request assistance, method for responding to equipment assistance request, and apparatuses
CN105182828B (en) * 2015-09-18 2018-09-07 深圳前海达闼科技有限公司 A kind of device request is assisted, response apparatus assists the method and apparatus of request
CN105553017A (en) * 2015-12-31 2016-05-04 小米科技有限责任公司 Robot charging system
CN105488892A (en) * 2016-01-04 2016-04-13 杭州亚美利嘉科技有限公司 Method and server for robot queuing management
CN105729491B (en) * 2016-04-01 2018-09-04 纳恩博(北京)科技有限公司 The execution method, apparatus and system of robot task
CN105729491A (en) * 2016-04-01 2016-07-06 纳恩博(北京)科技有限公司 Executing method, device and system for robot task
CN106777135A (en) * 2016-05-27 2017-05-31 中科鼎富(北京)科技发展有限公司 Service scheduling method, device and robot service system
CN106557362A (en) * 2016-11-15 2017-04-05 北京光年无限科技有限公司 It is applied to the task immigration method and system of intelligent robot
CN107223261A (en) * 2016-12-07 2017-09-29 深圳前海达闼云端智能科技有限公司 Man-machine complex decision method and apparatus
CN107077651A (en) * 2016-12-29 2017-08-18 深圳前海达闼云端智能科技有限公司 Robot cooperated method, device, robot and computer program product
CN107247460B (en) * 2017-06-01 2020-08-18 三峡大学 Cluster control method and system for machine bees
CN107247460A (en) * 2017-06-01 2017-10-13 三峡大学 A kind of cluster control method and system of machine honeybee
CN107450491A (en) * 2017-08-04 2017-12-08 杭州南江机器人股份有限公司 A kind of robot scheduling system, method, electronic equipment and storage medium
CN108171394A (en) * 2017-11-21 2018-06-15 中国科学院自动化研究所 Multi-robot Task Allocation based on hierachical structure and resource consolidation
WO2020103298A1 (en) * 2018-11-19 2020-05-28 炬星科技(深圳)有限公司 Distributed robot dispatch decision-making method, device and system, electronic device and storage medium

Also Published As

Publication number Publication date
CN102033536B (en) 2012-12-26

Similar Documents

Publication Publication Date Title
Bai et al. Optimal one-wafer cyclic scheduling and buffer space configuration for single-arm multicluster tools with linear topology
CN104461740B (en) A kind of cross-domain PC cluster resource polymerization and the method for distribution
Zhang et al. Model predictive control of autonomous mobility-on-demand systems
CN104040486B (en) Decouple PAAS resources, operation and scheduling
Rolón et al. Agent-based modeling and simulation of an autonomic manufacturing execution system
Xhafa et al. Meta-heuristics for grid scheduling problems
Ghiani et al. Real-time vehicle routing: Solution concepts, algorithms and parallel computing strategies
Maturana et al. Multi-agent mediator architecture for distributed manufacturing
Ouelhadj et al. A survey of dynamic scheduling in manufacturing systems
Krothapalli et al. Design of negotiation protocols for multi-agent manufacturing systems
CN103207814B (en) Managing and task scheduling system and dispatching method across cluster resource of a kind of decentration
CN101084680B (en) Method for managing resources in a platform for telecommunication service and/or network management, corresponding platform and computer program product therefor
CN101957780B (en) Resource state information-based grid task scheduling processor and grid task scheduling processing method
Hao et al. Implementing a hybrid simulation model for a Kanban-based material handling system
CN102681889B (en) Scheduling method of cloud computing open platform
Scholz-Reiter et al. Dynamic flexible flow shop problems—Scheduling heuristics vs. autonomous control
Liu et al. A survey on workflow management and scheduling in cloud computing
CN103426072B (en) The order processing system of a kind of high concurrent competition stock and disposal route thereof
CN104376365B (en) A kind of building method in the information system operation rule storehouse based on association rule mining
Decker et al. Quantitative modeling of complex environments
CN103944769B (en) Cluster resource system for unified management based on RPC agreements
CN102360310B (en) Multitask process monitoring method in distributed system environment
CN102193832B (en) Cloud computing resource scheduling method and applied system
CN101715003B (en) Load control optimization method of interoperable platform
Sharma et al. A survey of job scheduling and resource management in grid computing

Legal Events

Date Code Title Description
PB01 Publication
C06 Publication
SE01 Entry into force of request for substantive examination
C10 Entry into substantive examination
GR01 Patent grant
C14 Grant of patent or utility model
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20121226

Termination date: 20161222

CF01 Termination of patent right due to non-payment of annual fee