简介:TheAnalyticHierarchyProcessisapowerfultechniqueforgroupdecisionmaking.BoththeWeightedArithmeticMeanMethod(WAMM)andWeightedGeometricMeanMethod(WGMM)arethemostcommongrouppreferenceaggregationmethodsinAHP.InordertousetheWAMMandWGMM,onehastofindtheweightstobeassignedtothemembersofthegroup.Thisisoftenadifficulttask,especiallysoifthegroupislargeasinthecaseofpublicpolicydecisions.Thesesituationsneedanobjectivemethodtoderivemembers'weights.Butafewstudiesareavailableintheliterature.Basedonjudgementmatricesanderroranalyses,thispaperpresentstwopracticalandefficientmethodsforaddressingsuchsituations.Somenumericalexamplesarealsogiven.
简介:Ingroupdecisionmaking,acertaindegreeofconsensusisnecessarytoderiveameaningfulandvalidoutcome.ThispaperproposesaconsensusreachingmodelforagroupbyusingtheAnalyticHierarchyProcess(AHP).Itsupportspeopletoimprovetheirgroupconsensuslevelthroughanupdatingoftheirjudgments.Inthismodel,amoderatorsuggeststhemostdiscordantdecisionmakertoupdatehisjudgmentineachstep.Theproposedconsensusreachingmodelallowsdecisionmakerstoacceptorrejectthesuggestionfromthemoderator.Thismodelensuresthatthejudgmentupdatingiseffectiveandthefinalsolutionwillbeofacceptableconsistency.Finally,anumericalexampleisgiventoillustratethevalidityoftheproposedconsensusreachingmodel.
简介:LetG1andG2befinitedigraphs,bothwithvertexsetV.SupposethateachvertexvofVhasnonnegativeintegersf(v)andg(v)withf(v)≤g(v),andeacharceofG4hasnonnegativeintegersai(e)andbi(e)withai(e)≤bi(e),i=1,2.InthispaperwegiveanecessaryandsufficientconditionfortheexistenceofkarborescencesinG4coveringeachare(?)ofGiatleastai(e)andatmostbi(e)times,i=1,2,andsatisfyingtheconditionthatforeachvinVf(v)≤r1(v)=r2(v)≤g(v)wherer4(v)denotethenumberofthearborescencesinG?rootedatv.
简介:Inmultipleattributegroupdecisionmaking(MAGDM)problemsbasedonlinguisticinformation,thegranularitiesoflinguisticlabelsetsareusuallydifferentduetothedifferencesofthinkingmodesandhabitsamongdecisionmakers.Inordertodealwiththisinconvenience,thetransformationrelationshipsamongmultigranularlinguisticlabels(TRMLLs),whichareappliedtounifylinguisticlabelswithdifferentgranularitiesintoacertainlinguisticlabelsetwithfixedgranularity,arepresentedinthispaper.Furthermore,thereferencetablesaremadeaccordingtoTRMLLssothattheinterrelatedcalculationwillbelesscomplicated,andthemethodofhowtousethemisexplainedindetail.Atlength,theTRMLLsareillustratedthroughanapplicationexample.
简介:(GB)组买出现了并且在过去的十年演变为各种各样的形式。我们调查GB的一种不同形式,也就是消费者驱动的组买,一些消费者由此形成购买组一起访问店并且为折扣谈判。我们把这些消费者叫作不同于个别地访问店并且付出常规代价的常规消费者的GB消费者。由一个购买组,访问了一家店不得不做一个立即的决定在它的全部服务他们的需求。由它访问的第一家店拒绝了,购买组继续访问另外的店。在提供GB需求以后,店使用留下的股票为常规消费者服务。我们证明GB能是到采用合适的政策作为一台仪器利用它联系消费者的店的享受。店能在大多数情形以比常规价格低的价格提供组需求,但是设法仍然挣稳定的利润。常规消费者的存在在平衡形成上有微妙的效果,由加强在与微弱地做分类价格的消费者与组尺寸增加的GB谈判的店力量。而且,竞争的店能操作在购买组之间的相互作用并且当GB消费者说明市场底的小部分时,一起比一家独占者店挣更高全部的利润,竞争是强烈的。
简介:这篇论文问一个新问题:我们怎么能控制集体行为组织ofself的多代理人系统?我们试着由建议新观点called'SoftControl'回答这个问题,它把存在代理人的本地规则放在系统。我们由案例研究显示出软控制的可行性。考虑简单却典型的分布式的multi-agentmodel由Vicsek等求婚了。为鸟结队:每个代理人以一样的速度,但是与基于它的自己的标题的一般水准用一条本地规则被更新的不同标题和它的邻居的标题移动。这个模型的大多数研究关于self-organizedcollective行为,例如标题的同步。我们想要由软控制在这个组的集体行为(标题)干涉。一个指定方法是增加一个特殊代理人,叫了一个“党羽”,它能被我们控制,但是被另外的代理人当作一个平常的代理人。我们为党羽构造一条控制法律以便它能同步整个组到一个客观标题。这条控制法律被证明是有效经分解并且数字地。注意软控制与分布式的控制的途径不同。在分布式的系统干涉是一个自然方法。它可以在复杂系统的控制上拿出许多有趣的问题和挑战。
简介:1.IntroductionLetGbeafinitegroup.ForaCayleysubsetSofGnotcontainingtheidentityelement1,theCayley(di)graphX~Cay(G,S)ofGwithrespecttoSisdefinedasthedirectedgraphwithvertexsetV(X)=GandedgesetE(X)={(g,sg)IgEG,s6S}.IfS=S--’,thentheadjacencyrelationissymmet...
简介:AsubsetofSofthevertexsetofagraphGiscalledacyclicifthesubgraphitinducesinGcontainsnocycles.SiscalledanacyclicdominatingsetofGifitisbothacyclicanddominating.Theminimumcardinalityofanacyclicdominatingset,denotedbyγα(G),iscalledtheacyclicdominationnumberofG.S.M.Hedetniemietal.on2000introducedtheconceptofacyclicdominationandposedthefollowingopenproblem:Isγα(G)≤δ(G)foranygraphwhosediameteristwo?Inthispaper,wegiveacounterexamplewhichdisprovestheproblem.
简介:Thispaperdiscussestheharmonicproblemsincontrolsystemsfromtwoaspects:Oneistheharmoniccontrolamongdifferentsubsystems,andtheotheristheharmoniccontrolamongmultipleinputs.Someintrinsicproblemsinsuchsystemsarediscussed.Itispointedoutthatsomesubsystemsmustbeunstabletostabilizethewholeinterconnectedsystembyanexample.Especiallyfordiscrete-timemulti-inputsystems,anecessaryandsufficientconditionispresentedforthestrictdecreaseofthequadraticoptimalperformanceindexwiththecontrolinputextensions.Thisshowsanessentialdifferencebetweensingle-inputandmulti-inputcontrolsystems.Finally,somefutureresearchdirectionsarediscussedinharmoniccontrolofinterconnectedsystems,allocationofmulti-controlinputs,fault-tolerantcontrol,andfault-diagnosis.
简介:Inthispaper,theoptimalmaintenancepolicyisinvestigatedforasystemwithstochasticleadtimeandtwotypesoffailures.Thesystemhastwotypesoffailures,onetypeisrepairable,whentherepairablefailureoccurs,thesystemwillberepairedbyrepairman,andthesystemafterrepairisnot“asgoodasnew".Theothertypeoffailureisunrepairable,andwhentheunrepairablefailureoccursthesystemmustbereplacedbyanewandidenticalone.Thesparesystemforreplacementisavailableonlybyorder,andtheleadtimefordeliveringthesparesystemisstochastic.Thesuccessivesurvivaltimesofthesystemformastochasticallydecreasinggeometricprocess,theconsecutiverepairtimesafterfailuresofthesystemformarenewalprocess.Byusingtherenewalprocesstheoryandgeometricprocesstheory,theexplicitexpressionofthelong-runaveragecostperunittimeunderorderingpolicy(N-1)isderived,andthecorrespondingoptimalcanbefoundanalytically.Finally,thenumericalanalysesaregiven.
简介:Wefirstproposeaseriesofsimilaritymeasuresforintuitionisticfuzzyvalues(IFVs)basedontheintuitionisticfuzzyoperators(Atanassov1995).Theparametersintheproposedsimilaritymeasurescancontrolthedegreeofmembershipandthedegreeofnon-membershipofanIFV,whichcanreflectthedecisionmaker’sriskpreference.Moreover,wecanobtainsomeknownsimilaritymeasureswhensomefixedvaluesareassignedtotheparameters.Furthermore,weapplythesimilaritymeasurestoaggregateIFVsanddevelopsomeaggregationoperators,suchastheintuitionisticfuzzydependentaveragingoperatorandtheintuitionisticfuzzydependentgeometricoperator,whoseprominentcharacteristicisthattheassociatedweightsonlydependontheaggregatedintuitionisticfuzzyargumentsandcanrelievetheinfluenceofunfairargumentsontheaggregatedresults.Basedontheseaggregationoperators,wedevelopsomegroupdecisionmakingmethods,andfinallyextendourresultstointerval-valuedintuitionisticfuzzyenvironment.
简介:Farr-Gao算法是为减少的Gr的一个最先进的算法?bner有限的点的消失的理想基于,它作为一个内建的命令在枫树被实现了。这份报纸为它论述二维的改进那为计算减少的Gr采用预处理策略?与给定的点集合的塔子集联系的bner底。试验性的结果证明preprocessedFarr-Gao算法比古典的更有效。