CN107992999A - A kind of multiagent dispatching method towards personalized production environment - Google Patents
A kind of multiagent dispatching method towards personalized production environment Download PDFInfo
- Publication number
- CN107992999A CN107992999A CN201711200199.2A CN201711200199A CN107992999A CN 107992999 A CN107992999 A CN 107992999A CN 201711200199 A CN201711200199 A CN 201711200199A CN 107992999 A CN107992999 A CN 107992999A
- Authority
- CN
- China
- Prior art keywords
- autonomous agent
- user
- motion
- resource
- scheduling
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
Abstract
The present invention proposes a kind of multiagent dispatching method towards personalized production environment, Schilling resource side and each user are the scheduling main body with independent behaviour decision-making capability, and one initial schedule scheme of generation submits to the multiagent scheduling with of each user's autonomous agent at random by resource autonomous agent;Then resource autonomous agent is scheduled negotiation with user's autonomous agent by replacing motion, and when user's autonomous agent and resource autonomous agent fail to reach negotiation, there are will start alternately motion between user's autonomous agent of conflict to be scheduled negotiation;If resource autonomous agent and user's autonomous agent are consulted all to reach, consensus multiagent scheduling scheme is exported, otherwise remaining conflict is eliminated using preset arbitration rules and generates final multiagent scheduling scheme.It since this method has given full play to resource and the independence of each user, can fully meet the target call of each main body, improve the group's satisfaction degree of result scheduling.
Description
Technical field
The present invention relates to production system performance to optimize field, is specially a kind of multiagent tune towards personalized production environment
Degree method.
Background technology
Document " A hybrid algorithm based on a new neighborhood structure
evaluation method for job shop scheduling problem.Computers&Industrial
Engineering,2015,88:417-42 " discloses a kind of production system performance optimization method based on mixed scheduling algorithm.
This method carries out global optimization with particle cluster algorithm to the Maximal Makespan of scheduling first, then with variable neighborhood search algorithm into
Row local optimum, to minimize the Maximal Makespan of scheduling.This method obtain scheduling scheme due to Maximal Makespan more
It is small, thus the production efficiency of manufacturing enterprise can be improved, improve the utilization rate of resources of production and reduce production cost, but this method
It is only applicable to traditional large-scale production pattern.
Under large-scale production pattern, manufacturing enterprise usually by the demand merging treatment of different user without being distinguish between,
Then with related objective (such as Maximal Makespan) Optimized Operation of resource side, to reduce the manufacture cost of enterprise and improve production
Benefit.But as global market competition is increasingly sharpened, manufacturing informatization develops rapidly, traditional the operational mode of enterprise is
Through rapidly turning to personalized customization production from large-scale production pattern.Under personalized production environment, each user can have
Respective customized task and corresponding differentiation target, if still using the dispatch control method dominated with resource side at this time, just
It is difficult to the differentiation target call for meeting user, this is by the group's satisfaction degree for influencing user and reduces the industry competition of manufacturing enterprise
Power.
Resolve the scheduling controlling problem under personalized production environment, it is necessary to which the optimization aim of scheduling is by original main
Consider that the relevant target diversion of resource considers resource and the differentiation target of each user at the same time;The main body of scheduling decision is by resource side
Single main body be changed into resource side and multiagent that multi-user collectively constitutes.Thus a kind of new scheduling mould is just generated
Formula --- the multiagent towards personalized production environment is dispatched.
The content of the invention
The differentiation for being difficult to meet multi-user under personalized production environment for the existing dispatching method based on resource side
The problem of target requirement, the present invention propose a kind of multiagent dispatching method towards personalized production environment.This method Schilling provides
Source side and each user are the scheduling main body (i.e. resource autonomous agent and user's autonomous agent) with independent behaviour decision-making capability, and by
One initial schedule scheme of generation submits to each user's autonomous agent to start multiagent scheduling to resource autonomous agent at random;Then resource
By replacing motion, (user's autonomous agent receives the motion of resource autonomous agent to autonomous agent, then user's autonomous agent with user's autonomous agent
Change motion and feed back modification motion and give resource autonomous agent, then resource autonomous agent generates new motion and submits to each user certainly
Main body) negotiation is scheduled, and when user's autonomous agent and resource autonomous agent fail to reach negotiation, the user there are conflict is autonomous
Alternately motion will be started between body and be scheduled negotiation;If resource autonomous agent and user's autonomous agent are consulted all to reach, export
Consensus multiagent scheduling scheme, is otherwise eliminated using preset arbitration rules and remaining conflict and generate final more main
Body scheduling scheme.Since this method has given full play to resource and the independence of each user, it can fully meet the target of each main body
It is required that improve the group's satisfaction degree of result scheduling.
The technical scheme is that:
A kind of multiagent dispatching method towards personalized production environment, it is characterised in that:Comprise the following steps:
Step 1, multiagent scheduling is initialized:
(a) each user's autonomous agent is established with resource autonomous agent and communicated, to determine to participate in the autonomous agent set of multiagent scheduling
{A0,A1,A2,...,Ai,...,An, wherein A0Represent resource autonomous agent, Ai, i=1,2 ..., n represents each user's autonomous agent;
(b) resource autonomous agent A0With each user's autonomous agent Ai, i=1,2 ..., n communication, to determine adding for multiagent scheduling
Work task-set T={ T1,T2,...,Ti,...,Tn, wherein TiFor user's autonomous agent AiProcessing tasks;
(c) respective main body Ai, i=0,1,2 ..., n sets respective maximization regulation goal function Oi, i=0,1,
2 ..., n, and initialize respective consulting tactical S by formula (1)i;
In formula,For autonomous agent AiIt is t in negotiation timeiWhen motion target function value, β i are autonomous agent Ai's
Consulting tactical Dynamic gene,For autonomous agent AiDreamboat functional value,For autonomous agent AiAcceptable minimum target letter
Numerical value,For autonomous agent AiNegotiation deadline;
(d) each user's autonomous agent Ai, i=1,2 ..., n is initialized the acceptable conditions C of itself by formula (2)i, i=1,
2,...,n;
In formula, πi,jFor autonomous agent AiIt is received to come from autonomous agent AjScheduling motion, Oi(πi,j) it is autonomous agent AiDispatching
Motion πi,jUnder target function value;
(e) respective main body Ai, i=0,1,2 ..., n initializes respective warranty term N by formula (3)i, i=0,1,
2,...,n;
Step 2, multiagent scheduling is started:
Resource autonomous agent A0It is sky R={ } that constraint set is consulted in initialization, then using multiagent scheduling processing tasks collection T as tune
Object is spent, a scheduling scheme π is generated by the way of shown in formula (4)0, and submitted to the scheduling scheme as initial motion
Each user's autonomous agent Ai, i=1,2 ..., n is dispatched with starting multiagent;
Step 3, the negotiation information on-line study of Negotiation object is started:
When Negotiation object is autonomous agent AjWhen, in autonomous agent AjIts history proposal message collection is updated after submitting new motion, so
Afterwards using history motion and corresponding motion time as input, using the association shown in stochastic gradient descent method optimized-type (5)
Business's Policy model, with independent study body AjConsulting tactical Sj;The autonomous agent A obtained according to studyjApproximate consulting tacticalAsk
Derived from main body AjDreamboat functional value estimateThe estimate of acceptable minimum target functional valueAnd consult
The estimate of deadline
In formula,For autonomous agent AjIt is t in negotiation timejWhen motion target function value estimate,For autonomous agent
AjConsulting tactical Dynamic gene estimate, aj,bj,cjFor constant parameter;
Step 4, user's autonomous agent is consulted to dispatch with resource autonomous agent:User's autonomous agent passes through continuous with resource autonomous agent
Alternately motion and modification motion, it is final to obtain the consensus multiagent scheduling scheme of multiagent:
Step 4.1, user's autonomous agent independent behaviour decision-making:
(a) as user's autonomous agent AiReceive and come from resource autonomous agent A0Scheduling motion πi,0Afterwards, learn to provide by step 2
From main body A0Negotiation informationComprehensive resources autonomous agent A0Negotiation information X0And user is autonomous
Body AiNegotiation informationUser's autonomous agent AiIts consulting tactical Dynamic gene is changed using formula (6), so
Formula (1) is substituted into afterwards to adjust the consulting tactical S of itselfi
βi=βi+gi(X0,Xi) (6)
In formula, gi() is user's autonomous agent AiThe consulting tactical Tuning function of use;
(b) user's autonomous agent AiUsing the consulting tactical S after adjustmenti, renewal is from currently, negotiation time is tiWhen
The target function value of motionAdjust the acceptable conditions C of itselfiWith warranty term Ni, and negotiations process is controlled with this:If Ci
=1 notice resource autonomous agent A0Receive current motion and complete to consult, otherwise calculate Ni;If Ni=0, then notify resource certainly
Main body A0Refuse current motion and abandon consulting, otherwise user's autonomous agent Ai(c) is gone to step to continue to resource autonomous agent A0Motion;
(c) under conditions of meeting to consult constraint set R, user's autonomous agent AiBy formula (7), as a means of from main body A0Motion
πi,0Based on, obtained using local search algorithm and meet the desired value of itselfScheduling scheme, then using the program as
Feed back motion π0,iSubmit to resource autonomous agent A0;
Step 4.2, resource autonomous agent independent behaviour decision-making:
(a) resource autonomous agent waits and receives the feedback information of all user's autonomous agents for not terminating to consult:If user
Autonomous agent AiReceive motion, then resource autonomous agent A0Record user's autonomous agent AiComplete to consult and update to consult constraint set R;If
User's autonomous agent AiRefuse motion, then resource autonomous agent A0Identification and user's autonomous agent AiUser's autonomous agent A of conflictj, Ran Houzhuan
Step 5;If user's autonomous agent AiFeed back motion π0,i, then 2 study user's autonomous agent A are gone to stepiNegotiation information
(b) resource autonomous agent A0Judge whether multiagent scheduling is completed:Rule of judgment C is calculated by formula (8)0If C0=1,
Then multiagent scheduling is completed and goes to step 6, otherwise goes to step (c) and continues to consult;
In formula, BiFor user's autonomous agent AiNegotiation state mark:Bi=0 representative is abandoned consulting, Bi=1 representative continues to assist
Business, Bi=2, which represent completion, consults;
(c) comprehensive resources autonomous agent A0The negotiation information X of itself0And the negotiation of user's autonomous agent of each unfinished negotiation
Information { X1,X2... }, resource autonomous agent A0Its consulting tactical Dynamic gene is changed using formula (9), and substitutes into formula (1) renewal certainly
The consulting tactical S of body0;
β0=β0+g0(X0,X1,X2,...) (9)
In formula, g0() is resource autonomous agent A0The consulting tactical Tuning function of use;
(d) resource autonomous agent A0Using the consulting tactical S after adjustment0, renewal is from currently, negotiation time is t0When
The target function value of motionThe warranty term N of itself is adjusted by formula (3)0To control negotiations process:If Ni=0, then resource
Autonomous agent A0Continuation multiagent is abandoned to dispatch and go to step 6, otherwise under conditions of meeting to consult constraint set R, resource autonomous agent A0
Modification motion simultaneously makes new scheduling scheme π0Meet itself current target function valueThen by scheduling scheme π0Submit to institute
There is the user's autonomous agent for not completing negotiation;
Step 5, user's autonomous agent AiWith there are user's autonomous agent A of scheduling conflictjConsult scheduling:
Step 5.1, user's autonomous agent AjJudge whether to receive user's autonomous agent AiNegotiation request;If receiving negotiation,
5.2 are gone to step, otherwise user's autonomous agent AjRefusal and user's autonomous agent AiScheduling consult go to step 5.3;
Step 5.2, user's autonomous agent AiWith user's autonomous agent AjScheduling is consulted:
(a) user's autonomous agent AjIndependent behaviour decision-making:User's autonomous agent AjReception comes from user's autonomous agent AiScheduling motion
πj,iAfterwards, user's autonomous agent A is learnt by step 2iNegotiation informationAnd update the consulting tactical of itself
Sj, adjust the acceptable conditions C of itselfjWith warranty term Nj;If continuing motion, in the case where meeting to consult constraint set R,
With user's autonomous agent AiMotion πj,iBased on, user's autonomous agent AjObtained using local search algorithm and meet the desired value of itselfScheduling scheme, then using the program as feedback motion πi,jSubmit to user's autonomous agent Ai, otherwise user's autonomous agent AjKnot
Beam is consulted and goes to step 5.3;
(b) user's autonomous agent AiIndependent behaviour decision-making:User's autonomous agent AiReception comes from user's autonomous agent AjFeedback scheduling
Motion πi,jAfterwards, user's autonomous agent A is learnt by step 2jNegotiation informationAnd update the negotiation of itself
Tactful Si, adjust the acceptable conditions C of itselfiWith warranty term Ni;If continuing motion, in the situation for meeting to consult constraint set R
Under, with user's autonomous agent AjFeedback motion πi,jBased on, user's autonomous agent AiObtained using local search algorithm and meet itself
Desired valueScheduling scheme, then using scheme as new motion πj,iSubmit to user's autonomous agent Aj, otherwise user is autonomous
Body AiTerminate to consult and go to step 5.3;
Step 5.3, the step of going to step 4.2 (a) feeds back negotiation result to resource autonomous agent:If user's autonomous agent AiWith
User's autonomous agent AjReach negotiation, then the scheduling scheme π that will reach an agreementi,jAs user's autonomous agent AiWith user's autonomous agent Aj's
Resource autonomous agent A is submitted in common motion0;Otherwise user's autonomous agent AiNotify resource autonomous agent A0Abandon consulting and receive final
Multiagent scheduling scheme;
Step 6, if multiagent scheduling is completed, resource autonomous agent and the consensus tune of all user's autonomous agents are exported
Degree scheme π0,1,...,m;Otherwise the processing tasks for consulting not reach an agreement in addition to consulting constraint set R use preset arbitration rules
It is scheduled, generates final multiagent scheduling scheme π0,1,...,mAnd export.
Beneficial effect
The beneficial effects of the invention are as follows:Compared to traditional scheduler using resource as unique scheduling main body, present invention order money
Source and each user are provided simultaneously with the ability of Autonomous Scheduling decision-making, scheduling decision is turned by the traditional scheduler using resource as single main body
Multiagent to be participated in jointly with resource and each user is dispatched, so as to adapt to resource and the difference of each user under personalized production environment
Alienation target requirement;User and resource can fully meet the target of itself by the negotiation scheduling based on alternately motion, at the same time
Each main body can be fully understood by Negotiation object by on-line study, and then more reasonably adjust the consulting tactical of itself, make
Be conducive to meet own target, be beneficial to reach consensus scheduling decision;When can not directly reach one between user and resource
During cause, multiagent dispatching party can also be improved by the space of the fully excavation conflict resolution of the negotiation scheduling between user and user
Case;If each main body ultimately fails to reach consensus multiagent scheme, arbitration decisions will be final by set rule generation
Scheme, since whether the program can meet that the target of each main body is uncertain, this will encourage each main body to play an active part in negotiation
And decision-making is rationally scheduled, so as to reach consensus multiagent scheduling scheme as far as possible.It is in short, proposed by the present invention
Multiagent scheduling can play resource and the independence of each user, and fully meet the target call of each main body, improve result scheduling
Group's satisfaction degree.
The additional aspect and advantage of the present invention will be set forth in part in the description, and will partly become from the following description
Obtain substantially, or recognized by the practice of the present invention.
Brief description of the drawings
The above-mentioned and/or additional aspect and advantage of the present invention will become in the description from combination accompanying drawings below to embodiment
Substantially and it is readily appreciated that, wherein:
Fig. 1 is frame diagram of the present invention towards the multiagent dispatching method of personalized production environment.
Fig. 2 is the flow chart for consulting scheduling between the autonomous agent of the method for the present invention.
Fig. 3 is the flow chart of the autonomous agent independent behaviour decision-making of the method for the present invention.
Embodiment
The embodiment of the present invention is described below in detail, the embodiment is exemplary, it is intended to for explaining the present invention, and
It is not considered as limiting the invention.
With reference to Fig. 1-3, the present invention is comprised the following steps that towards the multiagent dispatching method of personalized production environment:
Step 1, multiagent scheduling is initialized:
(a) each user's autonomous agent establishes the autonomous agent set to communicate to determine to participate in multiagent scheduling with resource autonomous agent
{A0,A1,A2,...,Ai,...,An, wherein A0Represent resource autonomous agent, Ai, i=1,2 ..., n represents each user's autonomous agent;
(b) resource autonomous agent A0With each user's autonomous agent Ai, i=1,2 ..., n communications are with the processing of definite multiagent scheduling
Task-set T={ T1,T2,...,T1,...,Tn, wherein TiFor user's autonomous agent AiProcessing tasks;
(c) resource autonomous agent A0With each user's autonomous agent Ai, i=1,2 ..., n sets respective maximization regulation goal letter
Number Oi, i=0,1,2 ..., n simultaneously initializes respective consulting tactical S by formula (1)i;
In formula,For autonomous agent AiIt is t in negotiation timeiWhen motion target function value, βiFor autonomous agent AiAssociation
Business's Developing Tactics factor,For autonomous agent AiDreamboat functional value,For autonomous agent AiAcceptable minimum target function
Value,For autonomous agent AiNegotiation deadline;
(d) each user's autonomous agent Ai, i=1,2 ..., n initializes the acceptable conditions C of itself with formula (2)i, i=1,
2,...,n;
In formula, πi,jFor autonomous agent AiIt is received to come from autonomous agent AjScheduling motion, Oi(πi,j) it is autonomous agent AiDispatching
Motion πi,jUnder target function value;
(e) resource autonomous agent A0With each user's autonomous agent Ai, i=1,2 ..., n initializes respective negotiation bar with formula (3)
Part Ni, i=0,1,2 ..., n.
Step 2, multiagent scheduling is started:Resource autonomous agent A0It is sky R={ } that constraint set is consulted in initialization, then with more main
Body scheduling processing tasks integrate T as scheduler object, and a scheduling scheme π is generated by the way of shown in formula (4)0, and this is dispatched
Scheme submits to each user's autonomous agent A as initial motioni, i=1,2 ..., n is dispatched with starting multiagent.Its specific steps
For:
Step 3, the negotiation information on-line study of Negotiation object is started:When Negotiation object is autonomous agent AjWhen, in autonomous agent Aj
Its history proposal message collection is updated after submitting new motion, then using history motion and corresponding motion time as input,
Using the consulting tactical model shown in stochastic gradient descent method optimized-type (5), with independent study body AjConsulting tactical Sj;Root
The autonomous agent A obtained according to studyjApproximate consulting tacticalTry to achieve autonomous agent AjDreamboat functional value estimateCan
The estimate of the minimum target functional value of receivingAnd consult the estimate of deadlineIt is comprised the following steps that:
In formula,For autonomous agent AjIt is t in negotiation timejWhen motion target function value estimate,For autonomous agent
AjConsulting tactical Dynamic gene estimate, aj,bj,cjFor constant parameter.
(a) receive and come from Negotiation object AjPth proposal messageWherein,For Negotiation object AjPth is submitted to carry
CaseWhen negotiation time;
(b) the history proposal message collection of Negotiation object is updated:
(c) with object function collection { O1,O2,....,Oq,...,OmIn each object function assessment Negotiation object AjCarry
CaseAnd record the value { O of each object function1,p,O2,p,...,Oq,p,...,Om,p, wherein, Oq,p, q=1,2 ..., m is mesh
Scalar functions OqIn Negotiation object AjMotionUnder value;
(d) the target value data collection of Negotiation object history motion is updated:
(e) makeWithUsing stochastic gradient
Descent method, optimizes the consulting tactical model shown in formula (5), consults the consulting tactical of opponent with study;Then remember respectively
In each object function O under recordq, under q=1,2 ..., m, the model after studyLeast squares errorAnd the parameter of corresponding consulting tactical modelEstimate
(f) Negotiation object A is madejObject function OjForCorresponding object function Oθ,θ∈
{ 1,2 ..., m }, and make Negotiation object AjApproximate consulting tactical beThen simultaneous formula
WithTry to achieve Negotiation object AjDreamboat functional value estimateThe estimate of acceptable minimum target value
And consult the estimate of deadline
Step 4, user's autonomous agent is consulted to dispatch with resource autonomous agent:User's autonomous agent passes through continuous with resource autonomous agent
Alternately motion and modification motion, it is final to obtain the consensus multiagent scheduling scheme of multiagent.
Step 4.1, user's autonomous agent independent behaviour decision-making:
(a) as user's autonomous agent AiReceive and come from resource autonomous agent A0Scheduling motion πi,0Afterwards, learn to provide by step 2
From main body A0Negotiation informationComprehensive resources autonomous agent A0Negotiation information X0And user is autonomous
Body AiNegotiation informationUser's autonomous agent AiIts consulting tactical Dynamic gene is changed using formula (6), so
Formula (1) is substituted into afterwards to adjust the consulting tactical S of itselfi.Comprise the following steps that:
βi=βi+gi(X0,Xi) (6)
In formula, gi() is user's autonomous agent AiThe consulting tactical Tuning function of use.
As user's autonomous agent AiReceive and come from resource autonomous agent A0Scheduling motion πi,0Afterwards, resource autonomous agent A is recorded0Currently
Negotiation time t0;
By resource autonomous agent A0Negotiation time t0With scheduling motion πi,0Step 2 is submitted to learn and obtain resource autonomous agent
A0Negotiation information X0:Dreamboat functional valueAcceptable minimum target valueAnd consult deadline
User's autonomous agent AiThe negotiation information X of the resource autonomous agent obtained according to study0And user's autonomous agent AiItself
Negotiation information Xi, by formula (6) On-line accoun and change user's autonomous agent AiConsulting tactical Dynamic gene:
User's autonomous agent AiBy amended consulting tactical Dynamic gene βiSubstitution formula (1), adjusts the consulting tactical of itself
Si。
(b) user's autonomous agent AiUsing the consulting tactical S after adjustmenti, renewal is from currently, negotiation time is tiWhen
The target function value of motionAdjust the acceptable conditions C of itselfiWith warranty term Ni, and negotiations process is controlled with this:If Ci
=1 notice resource autonomous agent A0Receive current motion and complete to consult, otherwise calculate Ni;If Ni=0, then notify resource certainly
Main body A0Refuse current motion and abandon consulting, otherwise user's autonomous agent Ai(c) is gone to step to continue to resource autonomous agent A0Motion.
(c) under conditions of meeting to consult constraint set R, user's autonomous agent AiBy formula (7), as a means of from main body A0Motion
πi,0Based on, obtained using local search algorithm and meet the desired value of itselfScheduling scheme, then using the program as anti-
Present motion π0,iSubmit to resource autonomous agent A0。
Step 4.2, resource autonomous agent independent behaviour decision-making:
(a) resource autonomous agent waits and receives the feedback information of all user's autonomous agents for not terminating to consult:If user
Autonomous agent AiReceive motion, then resource autonomous agent A0Record user's autonomous agent AiComplete to consult and update to consult constraint set R;If
User's autonomous agent AiRefuse motion, then resource autonomous agent A0Identification and user's autonomous agent AiUser's autonomous agent A of conflicti, Ran Houzhuan
Step 5;If user's autonomous agent AiFeed back motion π0,i, then 2 study user's autonomous agent A are gone to stepiNegotiation informationComprise the following steps that:
Resource autonomous agent A0Wait user's autonomous agent A of all unfinished negotiationsiFeedback information:If user's autonomous agent
AiFeedback is Negotiation Decision Making information, then goes to step 2;If feedback is scheduling motion, 3 are gone to step;If user is autonomous
Body AiReceive current motion, then resource autonomous agent A0Record user's autonomous agent AiComplete to consult, while by user's autonomous agent AiPlus
Time interval where work task, which adds, consults constraint set R, positioned at the time interval other users autonomous agent consulted in constraint set R
Processing tasks must not take;If user's autonomous agent AiDisagree current motion and exit and resource autonomous agent A0Negotiation, that
Resource autonomous agent A0Will identification and user's autonomous agent AiThere is user's autonomous agent A of conflict on scheduling schemejAnd go to step 5.
Resource autonomous agent A0Reception comes from user's autonomous agent AiFeedback scheme π0,iAnd record user's autonomous agent AiAssociation
Business's time ti, and go to step 2 study user's autonomous agent AiNegotiation information Xi:Preferable target function valueAcceptable minimum
Desired valueAnd consult deadline
Resource autonomous agent A0Judge whether that all user's autonomous agents for not terminating to consult all have had been filed on feedback scheme:If
All have been filed on, go to step (b);Otherwise continue waiting for and receive the feedback information of user's autonomous agent.
(b) resource autonomous agent A0Judge whether multiagent scheduling is completed:Rule of judgment C is calculated by formula (8)0If C0=1,
Then multiagent scheduling is completed and goes to step 6, otherwise goes to step (c) and continues to consult.
In formula, BiFor user's autonomous agent AiNegotiation state mark:Bi=0 representative is abandoned consulting, Bi=1 representative continues to assist
Business, Bi=2, which represent completion, consults.
(c) comprehensive resources autonomous agent A0The negotiation information X of itself0And the negotiation of user's autonomous agent of each unfinished negotiation
Information { X1,X2... }, resource autonomous agent A0Its consulting tactical Dynamic gene is changed using formula (9), and substitutes into formula (1) renewal certainly
The consulting tactical S of body0.Concretely comprise the following steps:
β0=β0+g0(X0,X1,X2,...) (9)
In formula, g0() is resource autonomous agent A0The consulting tactical Tuning function of use.
Resource autonomous agent A0Each user's autonomous agent A obtained according to studyiNegotiation information Xi:Preferable target function valueAcceptable minimum target valueAnd consult deadlineWith resource autonomous agent A0The negotiation information X of itself0, pass through
Formula (9) On-line accoun simultaneously changes resource autonomous agent A0Consulting tactical Dynamic gene:
Resource autonomous agent A0By amended consulting tactical Dynamic gene β0Substitution formula (1), adjusts the consulting tactical of itself
S0。
(d) resource autonomous agent A0Using the consulting tactical S after adjustment0, renewal is from currently, negotiation time is t0When
The target function value of motionThe warranty term N of itself is adjusted by formula (3)0To control negotiations process:If Ni=0, then resource
Autonomous agent A0Continuation multiagent is abandoned to dispatch and go to step 6, otherwise under conditions of meeting to consult constraint set R, resource autonomous agent A0
Modification motion simultaneously makes new scheduling scheme π0Meet itself current target function valueThen by scheduling scheme π0Submit to institute
There is the user's autonomous agent for not completing negotiation.
Step 5, user's autonomous agent AiWith there are user's autonomous agent A of scheduling conflictjConsult scheduling:
Step 5.1, user's autonomous agent AjJudge whether to receive user's autonomous agent AiNegotiation request;If receiving negotiation,
5.2 are gone to step, otherwise user's autonomous agent AjRefusal and user's autonomous agent AiScheduling consult go to step 5.3;
Step 5.2, user's autonomous agent AiWith user's autonomous agent AjScheduling is consulted:
(a) user's autonomous agent AjIndependent behaviour decision-making:User's autonomous agent AjReception comes from user's autonomous agent AiScheduling motion
πj,iAfterwards, user's autonomous agent A is learnt by step 2iNegotiation informationAnd update the consulting tactical of itself
Sj, adjust the acceptable conditions C of itselfjWith warranty term Nj;If continuing motion, in the case where meeting to consult constraint set R,
With user's autonomous agent AiMotion πj,iBased on, user's autonomous agent AjObtained using local search algorithm and meet the desired value of itselfScheduling scheme, then using the program as feedback motion πi,jSubmit to user's autonomous agent Ai, otherwise user's autonomous agent AjKnot
Beam is consulted and goes to step 5.3;Concretely comprise the following steps:
As user's autonomous agent AjReceive and come from user's autonomous agent AiScheduling motion πj,iAfterwards, user's autonomous agent A is recordediCurrently
Negotiation time ti;
By user's autonomous agent AiNegotiation time tiWith scheduling motion πiStep 2 is submitted, learns and obtains user's autonomous agent
AiNegotiation information Xi:Dreamboat functional valueAcceptable minimum target valueAnd consult deadline
User's autonomous agent AjThe user's autonomous agent A obtained according to studyiNegotiation information XiAnd user's autonomous agent AjItself
Negotiation information Xj, On-line accoun simultaneously changes user's autonomous agent AiConsulting tactical Dynamic gene:
User's autonomous agent AjBy amended consulting tactical Dynamic gene βjSubstitution formula (1), adjusts the consulting tactical of itself
Sj.User's autonomous agent AjUsing the consulting tactical S after adjustmentj, renewal is from currently, negotiation time is tjWhen motion mesh
Offer of tender numerical valueAdjust the acceptable conditions C of itselfjWith warranty term Nj, and negotiations process is controlled with this:If Cj=1 logical
Know user's autonomous agent AiReceive current motion and complete to consult, otherwise calculate Nj;If Nj=0, then notify user's autonomous agent AiRefuse
Motion current absolutely simultaneously is abandoned consulting, otherwise user's autonomous agent AjContinue to user's autonomous agent AiMotion.
In the case where meeting to consult constraint set R, with user's autonomous agent AiMotion πj,iBased on, user's autonomous agent AiAdopt
Obtained with local search algorithm and meet itself current desired valueScheduling scheme, then using the program as feedback motion
πi,jSubmit to user's autonomous agent Ai。
(b) user's autonomous agent AiIndependent behaviour decision-making:User's autonomous agent AiReception comes from user's autonomous agent AjFeedback scheduling
Motion πi,jAfterwards, user's autonomous agent A is learnt by step 2jNegotiation informationAnd update the negotiation of itself
Tactful Si, adjust the acceptable conditions C of itselfiWith warranty term Ni;If continuing motion, in the situation for meeting to consult constraint set R
Under, with user's autonomous agent AjFeedback motion πi,jBased on, user's autonomous agent AiObtained using local search algorithm and meet itself
Desired valueScheduling scheme, then using scheme as new motion πj,iSubmit to user's autonomous agent Aj, otherwise user's autonomous agent
AiTerminate to consult and go to step 5.3;Concretely comprise the following steps:
As user's autonomous agent AiReceive and come from user's autonomous agent AjFeedback scheduling motion πi.jAfterwards, user's autonomous agent A is recordedj
Current negotiation time tj;
By user's autonomous agent AjNegotiation time tjWith scheduling motion πjStep 2 is submitted, learns and obtains user's autonomous agent
AjNegotiation information Xj:Dreamboat functional valueAcceptable minimum target valueAnd consult deadline
User's autonomous agent AiThe user's autonomous agent A obtained according to studyjNegotiation information XjAnd user's autonomous agent AiItself
Negotiation information Xi, On-line accoun simultaneously changes user's autonomous agent AiConsulting tactical Dynamic gene:
User's autonomous agent AiBy amended consulting tactical Dynamic gene βiSubstitution formula (1), adjusts the consulting tactical of itself
Si.User's autonomous agent AiUsing the consulting tactical S after adjustmenti, renewal is from currently, negotiation time is tiWhen motion mesh
Offer of tender numerical valueAdjust the acceptable conditions C of itselfiWith warranty term Ni, and negotiations process is controlled with this:If Ci=1 logical
Know user's autonomous agent AjReceive current motion and complete to consult, otherwise calculate Ni;If Ni=0, then notify user's autonomous agent AjRefuse
Motion current absolutely simultaneously is abandoned consulting, otherwise user's autonomous agent AiContinue to user's autonomous agent AjMotion.
In the case where meeting to consult constraint set R, with user's autonomous agent AjMotion πi,jBased on, user's autonomous agent AiAdopt
Obtained with local search algorithm and meet the desired value of itselfScheduling scheme, then using the program as feedback motion πj,iCarry
Give user's autonomous agent Aj。
Step 5.3, the step of going to step 4.2 (a) feeds back negotiation result to resource autonomous agent:If user's autonomous agent AiWith
User's autonomous agent AjReach negotiation, then the scheduling scheme π that will reach an agreementi,jAs user's autonomous agent AiWith user's autonomous agent Aj's
Resource autonomous agent A is submitted in common motion0;Otherwise user's autonomous agent AiNotify resource autonomous agent A0Abandon consulting and receive final
Multiagent scheduling scheme.
Step 6, if multiagent scheduling is completed, resource autonomous agent and the consensus tune of all user's autonomous agents are exported
Degree scheme π0,1,...,m;Otherwise the processing tasks do not reached an agreement to consulting the negotiation beyond constraint set R are advised using preset arbitration
Then it is scheduled, generates final multiagent scheduling scheme π0,1,...,mAnd export.
Although the embodiment of the present invention has been shown and described above, it is to be understood that above-described embodiment is example
Property, it is impossible to limitation of the present invention is interpreted as, those of ordinary skill in the art are not departing from the principle of the present invention and objective
In the case of above-described embodiment can be changed within the scope of the invention, change, replace and modification.
Claims (1)
- A kind of 1. multiagent dispatching method towards personalized production environment, it is characterised in that:Comprise the following steps:Step 1, multiagent scheduling is initialized:(a) each user's autonomous agent is established with resource autonomous agent and communicated, to determine to participate in the autonomous agent set { A of multiagent scheduling0, A1,A2,...,Ai,...,An, wherein A0Represent resource autonomous agent, Ai, i=1,2 ..., n represents each user's autonomous agent;(b) resource autonomous agent A0With each user's autonomous agent Ai, i=1,2 ..., n communication, to determine that the processing of multiagent scheduling is appointed Business collection T={ T1,T2,...,Ti,...,Tn, wherein TiFor user's autonomous agent AiProcessing tasks;(c) respective main body Ai, i=0,1,2 ..., n sets respective maximization regulation goal function Oi, i=0,1,2 ..., n, And initialize respective consulting tactical S by formula (1)i;<mrow> <msub> <mi>S</mi> <mi>i</mi> </msub> <mo>:</mo> <msubsup> <mi>O</mi> <mi>i</mi> <mi>n</mi> </msubsup> <mo>=</mo> <msubsup> <mi>O</mi> <mi>i</mi> <mi>g</mi> </msubsup> <mo>+</mo> <mrow> <mo>(</mo> <mn>1</mn> <mo>-</mo> <msup> <mrow> <mo>(</mo> <mfrac> <msub> <mi>t</mi> <mi>i</mi> </msub> <msubsup> <mi>t</mi> <mi>i</mi> <mi>d</mi> </msubsup> </mfrac> <mo>)</mo> </mrow> <msub> <mi>&beta;</mi> <mi>i</mi> </msub> </msup> <mo>)</mo> </mrow> <mrow> <mo>(</mo> <msubsup> <mi>O</mi> <mi>i</mi> <mi>e</mi> </msubsup> <mo>-</mo> <msubsup> <mi>O</mi> <mi>i</mi> <mi>g</mi> </msubsup> <mo>)</mo> </mrow> <mo>,</mo> <mi>i</mi> <mo>=</mo> <mn>0</mn> <mo>,</mo> <mn>1</mn> <mo>,</mo> <mn>2</mn> <mo>,</mo> <mo>...</mo> <mo>,</mo> <mi>n</mi> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>1</mn> <mo>)</mo> </mrow> </mrow>In formula,For autonomous agent AiIt is t in negotiation timeiWhen motion target function value, βiFor autonomous agent AiNegotiation plan Slightly Dynamic gene,For autonomous agent AiDreamboat functional value,For autonomous agent AiAcceptable minimum target functional value,For autonomous agent AiNegotiation deadline;(d) each user's autonomous agent Ai, i=1,2 ..., n is initialized the acceptable conditions C of itself by formula (2)i, i=1,2 ..., n;In formula, πi,jFor autonomous agent AiIt is received to come from autonomous agent AjScheduling motion, Oi(πi,j) it is autonomous agent AiIn scheduling motion πi,jUnder target function value;(e) respective main body Ai, i=0,1,2 ..., n initializes respective warranty term N by formula (3)i, i=0,1,2 ..., n;Step 2, multiagent scheduling is started:Resource autonomous agent A0It is sky R={ } that constraint set is consulted in initialization, then using multiagent scheduling processing tasks collection T as scheduling pair As generating a scheduling scheme π by the way of shown in formula (4)0, and submit to each use using the scheduling scheme as initial motion Family autonomous agent Ai, i=1,2 ..., n is dispatched with starting multiagent;Step 3, the negotiation information on-line study of Negotiation object is started:When Negotiation object is autonomous agent AjWhen, in autonomous agent AjUpdate its history proposal message collection after submitting new motion, then with History motion and corresponding motion time are used as input, using the negotiation plan shown in stochastic gradient descent method optimized-type (5) Slightly model, with independent study body AjConsulting tactical Sj;The autonomous agent A obtained according to studyjApproximate consulting tacticalTry to achieve certainly Main body AjDreamboat functional value estimateThe estimate of acceptable minimum target functional valueAnd consult to cut The only estimate of time<mrow> <msub> <mover> <mi>S</mi> <mo>^</mo> </mover> <mi>j</mi> </msub> <mo>:</mo> <msubsup> <mover> <mi>O</mi> <mo>^</mo> </mover> <mi>j</mi> <mi>n</mi> </msubsup> <mo>=</mo> <msub> <mi>a</mi> <mi>j</mi> </msub> <mo>-</mo> <msub> <mi>b</mi> <mi>j</mi> </msub> <msup> <mrow> <mo>(</mo> <mfrac> <msub> <mi>t</mi> <mi>j</mi> </msub> <msub> <mi>c</mi> <mi>j</mi> </msub> </mfrac> <mo>)</mo> </mrow> <msub> <mover> <mi>&beta;</mi> <mo>^</mo> </mover> <mi>j</mi> </msub> </msup> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>5</mn> <mo>)</mo> </mrow> </mrow>In formula,For autonomous agent AjIt is t in negotiation timejWhen motion target function value estimate,For autonomous agent Aj's The estimate of consulting tactical Dynamic gene, aj,bj,cjFor constant parameter;Step 4, user's autonomous agent is consulted to dispatch with resource autonomous agent:User's autonomous agent is with resource autonomous agent by constantly replacing Motion and modification motion, it is final to obtain the consensus multiagent scheduling scheme of multiagent:Step 4.1, user's autonomous agent independent behaviour decision-making:(a) as user's autonomous agent AiReceive and come from resource autonomous agent A0Scheduling motion πi,0Afterwards, by step 2 education resource certainly Main body A0Negotiation informationComprehensive resources autonomous agent A0Negotiation information X0And user's autonomous agent Ai Negotiation informationUser's autonomous agent AiIts consulting tactical Dynamic gene, Ran Houdai are changed using formula (6) Enter formula (1) to adjust the consulting tactical S of itselfiβi=βi+gi(X0,Xi) (6)In formula, gi() is user's autonomous agent AiThe consulting tactical Tuning function of use;(b) user's autonomous agent AiUsing the consulting tactical S after adjustmenti, renewal is from currently, negotiation time is tiWhen motion Target function valueAdjust the acceptable conditions C of itselfiWith warranty term Ni, and negotiations process is controlled with this:If Ci=1 Notify resource autonomous agent A0Receive current motion and complete to consult, otherwise calculate Ni;If Ni=0, then notify resource autonomous agent A0 Refuse current motion and abandon consulting, otherwise user's autonomous agent Ai(c) is gone to step to continue to resource autonomous agent A0Motion;(c) under conditions of meeting to consult constraint set R, user's autonomous agent AiBy formula (7), as a means of from main body A0Motion πi,0For Basis, is obtained using local search algorithm and meets the desired value of itselfScheduling scheme, then carried the program as feedback Case π0,iSubmit to resource autonomous agent A0;Step 4.2, resource autonomous agent independent behaviour decision-making:(a) resource autonomous agent waits and receives the feedback information of all user's autonomous agents for not terminating to consult:If user is autonomous Body AiReceive motion, then resource autonomous agent A0Record user's autonomous agent AiComplete to consult and update to consult constraint set R;If user Autonomous agent AiRefuse motion, then resource autonomous agent A0Identification and user's autonomous agent AiUser's autonomous agent A of conflictj, then go to step 5;If user's autonomous agent AiFeed back motion π0,i, then 2 study user's autonomous agent A are gone to stepiNegotiation information(b) resource autonomous agent A0Judge whether multiagent scheduling is completed:Rule of judgment C is calculated by formula (8)0If C0=1, then it is more Main body scheduling is completed and goes to step 6, otherwise goes to step (c) and continues to consult;In formula, BiFor user's autonomous agent AiNegotiation state mark:Bi=0 representative is abandoned consulting, Bi=1 representative continues to consult, Bi =2, which represent completion, consults;(c) comprehensive resources autonomous agent A0The negotiation information X of itself0And the negotiation information of user's autonomous agent of each unfinished negotiation {X1,X2... }, resource autonomous agent A0Its consulting tactical Dynamic gene is changed using formula (9), and substitutes into formula (1) and updates itself Consulting tactical S0;β0=β0+g0(X0,X1,X2,...) (9)In formula, g0() is resource autonomous agent A0The consulting tactical Tuning function of use;(d) resource autonomous agent A0Using the consulting tactical S after adjustment0, renewal is from currently, negotiation time is t0When motion Target function valueThe warranty term N of itself is adjusted by formula (3)0To control negotiations process:If Ni=0, then resource is autonomous Body A0Continuation multiagent is abandoned to dispatch and go to step 6, otherwise under conditions of meeting to consult constraint set R, resource autonomous agent A0Modification Motion simultaneously makes new scheduling scheme π0Meet itself current target function valueThen by scheduling scheme π0Submit to it is all not Complete the user's autonomous agent consulted;Step 5, user's autonomous agent AiWith there are user's autonomous agent A of scheduling conflictjConsult scheduling:Step 5.1, user's autonomous agent AjJudge whether to receive user's autonomous agent AiNegotiation request;If receiving negotiation, turn to walk Rapid 5.2, otherwise user's autonomous agent AjRefusal and user's autonomous agent AiScheduling consult go to step 5.3;Step 5.2, user's autonomous agent AiWith user's autonomous agent AjScheduling is consulted:(a) user's autonomous agent AjIndependent behaviour decision-making:User's autonomous agent AjReception comes from user's autonomous agent AiScheduling motion πj,i Afterwards, user's autonomous agent A is learnt by step 2iNegotiation informationAnd update the consulting tactical S of itselfj, Adjust the acceptable conditions C of itselfjWith warranty term Nj;If continuing motion, in the case where meeting to consult constraint set R, with Family autonomous agent AiMotion πj,iBased on, user's autonomous agent AjObtained using local search algorithm and meet the desired value of itself Scheduling scheme, then using the program as feedback motion πi,jSubmit to user's autonomous agent Ai, otherwise user's autonomous agent AjTerminate Consult and go to step 5.3;(b) user's autonomous agent AiIndependent behaviour decision-making:User's autonomous agent AiReception comes from user's autonomous agent AjFeedback scheduling motion πi,jAfterwards, user's autonomous agent A is learnt by step 2jNegotiation informationAnd update the consulting tactical of itself Si, adjust the acceptable conditions C of itselfiWith warranty term Ni;If continuing motion, in the case where meeting to consult constraint set R, With user's autonomous agent AjFeedback motion πi,jBased on, user's autonomous agent AiObtained using local search algorithm and meet the mesh of itself Scale valueScheduling scheme, then using scheme as new motion πj,iSubmit to user's autonomous agent Aj, otherwise user's autonomous agent Ai Terminate to consult and go to step 5.3;Step 5.3, the step of going to step 4.2 (a) feeds back negotiation result to resource autonomous agent:If user's autonomous agent AiAnd user Autonomous agent AjReach negotiation, then the scheduling scheme π that will reach an agreementi,jAs user's autonomous agent AiWith user's autonomous agent AjIt is common Resource autonomous agent A is submitted in motion0;Otherwise user's autonomous agent AiNotify resource autonomous agent A0Abandon consulting and receive final more Main body scheduling scheme;Step 6, if multiagent scheduling is completed, resource autonomous agent and the consensus dispatching party of all user's autonomous agents are exported Case π0,1,...,m;Otherwise the processing tasks do not reached an agreement to consulting constraint set R to consult in addition are carried out using preset arbitration rules Scheduling, generates final multiagent scheduling scheme π0,1,...,mAnd export.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711200199.2A CN107992999B (en) | 2017-11-27 | 2017-11-27 | Multi-subject scheduling method oriented to personalized production environment |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711200199.2A CN107992999B (en) | 2017-11-27 | 2017-11-27 | Multi-subject scheduling method oriented to personalized production environment |
Publications (2)
Publication Number | Publication Date |
---|---|
CN107992999A true CN107992999A (en) | 2018-05-04 |
CN107992999B CN107992999B (en) | 2021-06-11 |
Family
ID=62032145
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201711200199.2A Active CN107992999B (en) | 2017-11-27 | 2017-11-27 | Multi-subject scheduling method oriented to personalized production environment |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107992999B (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112101712A (en) * | 2020-08-04 | 2020-12-18 | 西北工业大学 | Parallel machine multi-agent auction negotiation scheduling method |
CN112529313A (en) * | 2020-12-17 | 2021-03-19 | 中国航空综合技术研究所 | Intelligent human-machine engineering design optimization method based on negotiation strategy |
CN112862312A (en) * | 2021-02-07 | 2021-05-28 | 山东大学 | Manufacturing service resource dynamic scheduling method and system based on random online algorithm |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1302403A (en) * | 1998-05-22 | 2001-07-04 | 弗兰普顿·E·埃利斯三世 | Global network computer |
US7139999B2 (en) * | 1999-08-31 | 2006-11-21 | Accenture Llp | Development architecture framework |
CN104023316A (en) * | 2013-03-01 | 2014-09-03 | 华为技术有限公司 | Multicast information transmission method and device |
CN106375086A (en) * | 2016-08-27 | 2017-02-01 | 张春萍 | Big data-based internet teaching system running method |
-
2017
- 2017-11-27 CN CN201711200199.2A patent/CN107992999B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1302403A (en) * | 1998-05-22 | 2001-07-04 | 弗兰普顿·E·埃利斯三世 | Global network computer |
US7139999B2 (en) * | 1999-08-31 | 2006-11-21 | Accenture Llp | Development architecture framework |
CN104023316A (en) * | 2013-03-01 | 2014-09-03 | 华为技术有限公司 | Multicast information transmission method and device |
CN106375086A (en) * | 2016-08-27 | 2017-02-01 | 张春萍 | Big data-based internet teaching system running method |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112101712A (en) * | 2020-08-04 | 2020-12-18 | 西北工业大学 | Parallel machine multi-agent auction negotiation scheduling method |
CN112101712B (en) * | 2020-08-04 | 2023-06-09 | 西北工业大学 | Parallel machine multi-agent auction negotiation scheduling method |
CN112529313A (en) * | 2020-12-17 | 2021-03-19 | 中国航空综合技术研究所 | Intelligent human-machine engineering design optimization method based on negotiation strategy |
CN112529313B (en) * | 2020-12-17 | 2022-12-09 | 中国航空综合技术研究所 | Intelligent human-machine engineering design optimization method based on negotiation strategy |
CN112862312A (en) * | 2021-02-07 | 2021-05-28 | 山东大学 | Manufacturing service resource dynamic scheduling method and system based on random online algorithm |
Also Published As
Publication number | Publication date |
---|---|
CN107992999B (en) | 2021-06-11 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Chang et al. | Optimisation of distributed manufacturing flexible job shop scheduling by using hybrid genetic algorithms | |
CN107992999A (en) | A kind of multiagent dispatching method towards personalized production environment | |
CN111756653B (en) | Multi-coflow scheduling method based on deep reinforcement learning of graph neural network | |
CN109270904A (en) | A kind of flexible job shop batch dynamic dispatching optimization method | |
US6647300B1 (en) | Bidding partner cache for autonomous cooperative control system | |
CN104537503B (en) | Data processing method and system | |
CN102346882B (en) | Bilayer planning method for intelligent scheduling for steelmaking-continuous welding production | |
CN111199360A (en) | Unmanned aerial vehicle task allocation planning method | |
US11846926B2 (en) | Negotiation-based method and system for coordinating distributed MES order management | |
CN105974891B (en) | A kind of mold production process self-adaptation control method based on dynamic billboard | |
CN108027608A (en) | For the unlatching of the processing step of process object | |
Soltero et al. | Using Kaizen to reduce waste and prevent pollution. | |
US20070039004A1 (en) | Decentralized coordination of resource usage in multi-agent systems | |
CN103810562A (en) | Dynamic body synergetic method for supply chain system | |
CN114503038A (en) | Method and apparatus for self-learning manufacturing schedule for flexible manufacturing system using state matrix | |
Piya et al. | Negotiation based decision support system for order acceptance | |
CN102402734A (en) | Machining and assembly alternative mixed scheduling method under flexible path | |
Zaman et al. | Resource constrained project scheduling with dynamic disruption recovery | |
Zhang et al. | Optimal production scheduling with multi-round information interaction for demander-dominated decentralized scheduling problem | |
CN116596672A (en) | Intelligent contract task distribution platform, method, system and medium based on block chain | |
CN106900007A (en) | Towards the complex task distribution method of the wireless sensor network for overlapping sub-clustering | |
Skobelev et al. | Practical approach and multi-agent platform for designing real time adaptive scheduling systems | |
CN104036336A (en) | Dynamic body synergy method for supply chain systems | |
Beham et al. | A genetic programming approach to solve scheduling problems with parallel simulation | |
López-Ortega et al. | Intelligent and collaborative Multi-Agent System to generate and schedule production orders |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |