基于数据挖掘的电力物资评标系统的研究与应用

分类号TP391密级UDC硕士学位论文基于数据挖掘的电力物资评标系统的研究与应用学位申请人：陈晓学科专业：计算机应用技术指导教师：蒋廷耀教授二○一二年五月ADissertationSubmittedinPartialFulfillmentoftheRequirementsfortheDegreeofMasterofScienceinEngineeringTheResearchandApplicationofElectricitySuppliesBidEvaluationSystemBasedonDataMiningGraduateStudent:ChenXiaoMajor:ComputerAppliedTechnologySupervisor:Prof.JiangTingyaoChinaThreeGorgesUniversityYichang,443002,P.R.ChinaMay,2012三峡大学硕士学位论文三峡大学学位论文原创性声明本人郑重声明：所呈交的学位论文，是本人在导师的指导下，独立进行研究工作所取得的成果，除文中已经注明引用的内容外，本论文不含任何其他个人或集体已经发表或撰写过的作品成果。对本文的研究做出重要贡献的个人和集体均已在文中以明确方式标明，本人完全意识到本声明的法律后果由本人承担。学位论文作者签名：日期：I三峡大学硕士学位论文内容摘要随着我国电力工程的快速发展，电力工程中的招投标活动影响电力行业的发展，电力物资的评标是招投标环节的核心内容。现有的评标方案中，对于小宗物资的评标比较完善，但涉及到大宗物资的评标时，则其评标方案存在评标专家主观意识影响较大、对各种影响因素未全面考虑，找出各个因素对评标结果的影响程度、计算过程存在单一化，评标结果未采取相关横向对照等问题。因此设计一套完整，公平的评标方法是当前的研究热点。本文根据电力物资的评标活动的特殊性，首先阐述了电力物资招投标活动的流程，然后重点设计了电力物资的评标流程。在评标流程中，评标方法是最核心的部分，文章首先介绍了当前常用的评标方法，然后重点介绍了综合评分法，并详述了其算法流程，提出了方法存在的问题。针对综合评分法存在的问题，提出了一种改进的基于层次分析法的评标模型，并对算法可靠性与可行性做了分析。针对历史评标数据，采用数据挖掘技术分析历史评标数据的内在规律，提出了一种基于BP神经网络的评标决策支持模型，通过该模型，提出了投标人信誉得分的概念，然后将决策支持系统与改进的基于AHP的综合评分法结合起来，提出了信誉权重的概念，提出了数据挖掘的评标模型。最后，通过实际评标数据，分别对几种不同的评标算法模型进行了验证与结果分析，总结了评标模型的优点，同时也指出了当前的评标模型需要进一步研究的方向。关键词：综合评分法层次分析法数据挖掘BP神经网络II三峡大学硕士学位论文AbstractWiththerapiddevelopmentofourcountry’selectricpowerengineering,electricpowerengineeringbiddingactivitiesaffectthedevelopmentofelectricpowerindustry.Thebidevaluationofpowersuppliesisthecorecontentofthebiddingprocess.Intheexistingofbidevaluationschemes,thebidevaluationofsmallgoodsisveryperfectwhilethebidevaluationofbulkgoodshassomeproblems,forexample:theinfluenceofvariousfactorsforinstance,theexperts’subjectiveconsciousnessinfluence,onthebidevaluationarenotfullyconsidered.Thecalculationprocessofevaluationresultsissinglethatthereisnohorizontalcomparisonbetweenfactors.Therefore,thecurrentresearchfocusesonthedesignsofasetofcompleteandfairbiddingevaluationmethods.Inthispaper,accordingtotheparticularityofthebidevaluationactivitiesofelectricpowersupplies,first,thepoweroftenderingandbiddingactivitiesoftheprocessiselaborated,andthenpowersuppliesbiddingprocessisfocusedontherated.Intheprocessofbidevaluation,bidevaluationmethodistheessentialpart.Inthispaper,thecurrentcommonlyusedevaluationmethodsarefirstintroduced,andthenthecomprehensivescoringmethodfocusedonthedescriptionoftheprocessofalgorithm.Inthissense,theproblemofthismethodisputforward.Basedontheproblemsofcomprehensivescoringmethod,akindofimprovedAHPbasedevaluationmodelisputforward,andthereliabilityandfeasibilityofalgorithmisanalyzed.Accordingtotheanalysisofhistoricalevaluationdataandtheapplicationofdataminingtechnology,theinherentlawsofhistoricalevaluationdatacanbeanalyzed.ThenakindofbiddingevaluationdecisionsupportmodelbasedontheBPneuralnetworkisproposed.Throughthismodel,theconceptofcreditscorecanbeputforward.FinallycombinethebiddingevaluationdecisionsupportmodelbasedontheBPneuralnetworkandtheimprovedcomprehensivescoringmethodbasedonAHPtogether,theconceptofcreditweightandtheevaluationmodelofdataminingareputforward.Finally,throughtheactualevaluationdata,severaldifferentkindsofevaluationalgorithmmodelsarevalidatedandtheresultisanalyzed.Thenmeritoftheevaluationmodelissummarized,andthefurtherresearchdirectionofthecurrentevaluationmodelispointedout.Keywords:ThecomprehensivescoringmethodAnalytichierarchyprocessDataminingBPneuralnetworkIII三峡大学硕士学位论文目录引言...................................................................................................................11绪论................................................................................................................31.1研究背景..........................................................................................31.2论文的目的和意义..........................................................................31.3论文的主要工作..............................................................................41.4论文的组织结构..............................................................................52电力物资招标评标现状................................................................................62.1电力物资招投标流程.....................................................................62.2评标................................................................................................92.3电力物资评标方法.......................................................................122.4本章小结.......................................................................................143评标方法的研究与改进..............................................................................153.1综合评分法...................................................................................153.2层次分析法(AHP)..........................................................................163.3综合评分法的改进.......................................................................173.4实例计算.......................................................................................193.5算法评价.......................................................................................213.6本章小结.......................................................................