基于報童模型的市場需求預(yù)測_第1頁
基于報童模型的市場需求預(yù)測_第2頁
基于報童模型的市場需求預(yù)測_第3頁
基于報童模型的市場需求預(yù)測_第4頁
基于報童模型的市場需求預(yù)測_第5頁
已閱讀5頁,還剩87頁未讀 繼續(xù)免費(fèi)閱讀

下載本文檔

版權(quán)說明:本文檔由用戶提供并上傳,收益歸屬內(nèi)容提供方,若內(nèi)容存在侵權(quán),請進(jìn)行舉報或認(rèn)領(lǐng)

文檔簡介

#參考文獻(xiàn)蘇欣,劉光潔?帶有預(yù)算費(fèi)用約束的多地點(diǎn)報童模型[J].長春工程學(xué)院學(xué)報(自然科學(xué)版).2003年,4(2):23-28.LauH.S.Thenewsboyproblemunderalternativeoptimizationobjectives[J].JournaloftheOperationsResearchSociety,1980,31:525-535.AgrawalV,SeshadriS.Impactofuncertaintyandriskaversiononpriceandorderquantityinthenewsvendorproblem[J].Manufacturing&ServiceOperationsManagement,2000,2:410-423.蘇欣,林正華,楊麗.一次訂購季節(jié)性銷售的一種擴(kuò)展報童模型[J].吉林大學(xué)學(xué)報(理學(xué)版),2003,41(3):314-318.高尚?價格有折扣的報童問題[J].華東船舶工業(yè)學(xué)院學(xué)報(自然科學(xué)版),2001年8月,15(4):65-68.姚洪義,高云靜?價格競爭下基于報童問題的需求模型分析[J].重慶大學(xué)學(xué)報(自然科學(xué)版),2007年11月,30(11):144-147.HolmstromB.Moralhazardandobservability[J]BellJournalofEconomics1979,10:74291.蔡清波,魯其輝,朱道立預(yù)測精度隨時間變化的報童模型分析[J].預(yù)測,2003年,5(10):43-46.李明琨,汪凱仁,方芳?基于時間因素的報童問題理論方法研究[J].系統(tǒng)工程理論方法應(yīng)用,2003年6月,12(2):146-152.吳鵬?考慮回收再制造的報童模型擴(kuò)展[J]?清華大學(xué)學(xué)報(哲學(xué)社會科學(xué)版),2006年,1(21):71-76.宋海濤,林正華,蘇欣帶有二次訂購和二次銷售的報童問題[J].經(jīng)濟(jì)數(shù)學(xué),2003年3月,20(1):73-80.劉麗華,曾玲?可追加訂購報童問題的模糊機(jī)會約束規(guī)劃模型[J].汕頭大學(xué)學(xué)報(自然科學(xué)版),2007年2月,22(1):2-6.于春云,趙希男,彭艷東,潘德惠?模糊隨機(jī)需求模式下的擴(kuò)展報童模型與求解算法[J].系統(tǒng)工程,2006年9月,24(9):103-107.宋海濤,王秋月?三角分布下可追加訂購的報童問題最優(yōu)解[J].內(nèi)蒙古民族大學(xué)學(xué)報(自然科學(xué)版),2003年8月,18(4):289-292.宋海濤,林正華.二次降價銷售的報童問題[J].吉林大學(xué)學(xué)報(理學(xué)版),2004年10月,42(4):42-47.鮑芳?基于數(shù)據(jù)挖掘方法的產(chǎn)品市場需求預(yù)測[J].決策參考,2006年,6(12):44-50.致謝在論文完成之際,謹(jǐn)向多年來關(guān)心、支持我的人們致以最誠摯的謝意!首先感謝我的導(dǎo)師陳培金老師和劉合翔老師在近半年來對我畢業(yè)論文的悉心指導(dǎo)!從選題到考題報告,從文獻(xiàn)綜述到中期報告,從論文正文到以后建議和指正,都體現(xiàn)了兩位老師的細(xì)心負(fù)責(zé)的優(yōu)良教風(fēng)和淵博的知識,讓我受益匪淺。感謝家人和課任教師學(xué)校對我這么多年來的悉心栽培,沒有你們也就沒有今天的我,學(xué)校讓我感受到了家的溫暖,同學(xué)讓我感受到了家人般的關(guān)愛,在大學(xué)4年中也提供了很大的幫助。衷心地向你們說聲謝謝!最后衷心祝愿所有的同學(xué)都能工作順利,所有幫過我的老師能多出成果、心想事成,希望學(xué)校能更上一個臺階,忠心感謝在百忙之中抽出寶貴時間對此論文進(jìn)行評閱與審議的老師們。浙江林學(xué)院本(專)科生畢業(yè)設(shè)計(論文)任務(wù)書設(shè)計(論文)題目:基于報童模型的市場需求預(yù)測學(xué)院名稱:信息工程學(xué)院專業(yè)班級:信息043學(xué)生姓名:鄭歡歡學(xué)號:200405021307指導(dǎo)教師:陳培金、劉合翔學(xué)科負(fù)責(zé)人:1.設(shè)計(論文)的主要任務(wù)和目標(biāo)目標(biāo):通過查閱大量的文獻(xiàn)資料,自主學(xué)習(xí)并掌握豐富、全面的有關(guān)報童模型的相關(guān)知識;了解報童模型產(chǎn)生的原因、發(fā)展?fàn)顩r以及研究現(xiàn)狀;要求深刻理解并掌握報童模型的概念及其內(nèi)涵,了解報童模型適用的研究環(huán)境;要求能熟練的運(yùn)用報童模型針對企業(yè)的生產(chǎn)困境進(jìn)行分析研究,提出解決方法;要求能運(yùn)用報童模型對企業(yè)進(jìn)行市場需求預(yù)測,并要求能計算相應(yīng)的績效指標(biāo)。主要任務(wù):閱讀大量有關(guān)報童模型的文獻(xiàn);掌握報童模型的概念及其內(nèi)涵;針對課題研究的問題,對溫州鵬昌皮革有限公司進(jìn)行實(shí)地調(diào)查;收集論論研究所需要的相關(guān)數(shù)據(jù);運(yùn)用報童模型對所收集到的數(shù)據(jù)進(jìn)行數(shù)據(jù)分析;根據(jù)數(shù)據(jù)分析的結(jié)果,對企業(yè)的生產(chǎn)困境提出解決方法;說明進(jìn)行論文研究的意義;完成論文;2.設(shè)計(論文)的主要內(nèi)容運(yùn)用報童模型對企業(yè)進(jìn)行全面的數(shù)據(jù)分析,根據(jù)企業(yè)的歷史銷售狀況和目前企業(yè)的生產(chǎn)狀況確定企業(yè)對產(chǎn)品的最優(yōu)訂購量。以求達(dá)到減少原材料無謂的浪費(fèi)、降低生產(chǎn)成本;減少庫存量,提高倉庫利用效率;提高訂單完成率,獲得最佳效益的目的。論文應(yīng)包括一下幾方面的內(nèi)容:運(yùn)用報童模型預(yù)測企業(yè)利潤最大化的最佳訂購量;運(yùn)用報童模型預(yù)測在達(dá)到一定目標(biāo)訂單完成率要求下的最佳訂購量;運(yùn)用報童模型預(yù)測在達(dá)到一定目標(biāo)存貨滿足率要求下的最佳訂購量;數(shù)據(jù)分析實(shí)時展現(xiàn),幫助管理層作出正確的決策分析;對溫州鵬昌皮革有限公司運(yùn)用報童模型進(jìn)行預(yù)測提出自己的觀點(diǎn)和意見;對溫州鵬昌皮革有限公司的生產(chǎn)狀況進(jìn)行分析及預(yù)測。3.設(shè)計(論文)的基本要求報童模型的概念認(rèn)識和熟悉市場需求預(yù)測的概念認(rèn)識和熟悉對溫州鵬昌皮革有限公司的生產(chǎn)狀況進(jìn)行分析及預(yù)測分析各項績效指標(biāo)對溫州鵬昌皮革有限公司運(yùn)用報童模型進(jìn)行預(yù)測提出自己的觀點(diǎn)和意見4.主要參考文獻(xiàn)蘇欣,劉光潔?帶有預(yù)算費(fèi)用約束的多地點(diǎn)報童模型[J].長春工程學(xué)院學(xué)報(自然科學(xué)版).2003年,4(2):23-28.LauH.S.Thenewsboyproblemunderalternativeoptimizationobjectives[J].JournaloftheOperationsResearchSociety,1980,31:525-535.AgrawalV,SeshadriS.Impactofuncertaintyandriskaversiononpriceandorderquantityinthenewsvendorproblem[J].Manufacturing&ServiceOperationsManagement,2000,2:410-423.蘇欣,林正華,楊麗.一次訂購季節(jié)性銷售的一種擴(kuò)展報童模型[J].吉林大學(xué)學(xué)報(理學(xué)版),2003,41(3):314-318.高尚?價格有折扣的報童問題[J].華東船舶工業(yè)學(xué)院學(xué)報(自然科學(xué)版),2001年8月,15(4):65-68.姚洪義,高云靜?價格競爭下基于報童問題的需求模型分析[J].重慶大學(xué)學(xué)報(自然科學(xué)版),2007年11月,30(11):144-147.HolmstromB.Moralhazardandobservability[J]BellJournalofEconomics1979,10:74291.蔡清波,魯其輝,朱道立預(yù)測精度隨時間變化的報童模型分析[J].預(yù)測,2003年,5(10):43-46.李明琨,汪凱仁,方芳?基于時間因素的報童問題理論方法研究[J].系統(tǒng)工程理論方法應(yīng)用,2003年6月,12(2):146-152.吳鵬?考慮回收再制造的報童模型擴(kuò)展[J]?清華大學(xué)學(xué)報(哲學(xué)社會科學(xué)版),2006年,1(21):71-76.宋海濤,林正華,蘇欣帶有二次訂購和二次銷售的報童問題[J].經(jīng)濟(jì)數(shù)學(xué),2003年3月,20(1):73-80.文U麗華,曾玲?可追加訂購報童問題的模糊機(jī)會約束規(guī)劃模型[J].汕頭大學(xué)學(xué)報(自然科學(xué)版),2007年2月,22(1):2-6.于春云,趙希男,彭艷東,潘德惠?模糊隨機(jī)需求模式下的擴(kuò)展報童模型與求解算法[J].系統(tǒng)工程,2006年9月,24(9):103-107.宋海濤,王秋月?三角分布下可追加訂購的報童問題最優(yōu)解[J].內(nèi)蒙古民族大學(xué)學(xué)報(自然科學(xué)版),2003年8月,18(4):289-292.宋海濤,林正華.二次降價銷售的報童問題[J].吉林大學(xué)學(xué)報(理學(xué)版),2004年10月,42(4):42-47.鮑芳?基于數(shù)據(jù)挖掘方法的產(chǎn)品市場需求預(yù)測[J].決策參考,2006年,6(12):44-50.5.進(jìn)度安排設(shè)計(論文)各階段名稱起止日期1查閱資料、調(diào)查研究、資料收集、整理、外文翻譯2007.07——-2007.092寫出開題報告、文件綜述,聽取導(dǎo)師意見,進(jìn)行修改2007.10——-2008.013材料補(bǔ)充、分析,完成論文初稿2008.02——-2008.034聽取導(dǎo)師意見,論文修改充實(shí)2008.04——-2008.055論文答辯2008.5注:一式三份,學(xué)院、指導(dǎo)教師、學(xué)生各一份,由指導(dǎo)教師填寫。英語原文1:DemandForecastingTheImportanceofDemandForecastingForecastingproductdemandiscrucialtoanysupplier,manufacturer,orretailer.Forecastsoffuturedemandwilldeterminethequantitiesthatshouldbepurchased,produced,andshipped.Demandforecastsarenecessarysincethebasicoperationsprocess,movingfromthesuppliers'rawmaterialstofinishedgoodsinthecustomers'hands,takestime.Mostfirmscannotsimplywaitfordemandtoemergeandthenreacttoit.Instead,theymustanticipateandplanforfuturedemandsothattheycanreactimmediatelytocustomerordersastheyoccur.Inotherwords,mostmanufacturers"maketostock"ratherthan"maketoorder"一theyplanaheadandthendeployinventoriesoffinishedgoodsintofieldlocations.Thus,onceacustomerordermaterializes,itcanbefulfilledimmediately一sincemostcustomersarenotwillingtowaitthetimeitwouldtaketoactuallyprocesstheirorderthroughoutthesupplychainandmaketheproductbasedontheirorder.Anordercyclecouldtakeweeksormonthstogobackthroughpartsuppliersandsub-assemblers,throughmanufactureoftheproduct,andthroughtotheeventualshipmentoftheordertothecustomer.Firmsthatofferrapiddeliverytotheircustomerswilltendtoforceallcompetitorsinthemarkettokeepfinishedgoodinventoriesinordertoprovidefastordercycletimes.Asaresult,virtuallyeveryorganizationinvolvedneedstomanufactureoratleastorderpartsbasedonaforecastoffuturedemand.Theabilitytoaccuratelyforecastdemandalsoaffordsthefirmopportunitiestocontrolcoststhroughlevelingitsproductionquantities,rationalizingitstransportation,andgenerallyplanningforefficientlogisticsoperations.Ingeneralpractice,accuratedemandforecastsleadtoefficientoperationsandhighlevelsofcustomerservice,whileinaccurateforecastswillinevitablyleadtoinefficient,highcostoperationsand/orpoorlevelsofcustomerservice.Inmanysupplychains,themostimportantactionwecantaketoimprovetheefficiencyandeffectivenessofthelogisticsprocessistoimprovethequalityofthedemandforecastsGeneralApproachestoForecastingAllfirmsforecastdemand,butitwouldbedifficulttofindanytwofirmsthatforecastdemandinexactlythesameway.Overthelastfewdecades,manydifferentforecastingtechniqueshavebeendevelopedinanumberofdifferentapplicationareas,includingengineeringandeconomics.Manysuchprocedureshavebeenappliedtothepracticalproblemofforecastingdemandinalogisticssystem,withvaryingdegreesofsuccess.Mostcommercialsoftwarepackagesthatsupportdemandforecastinginalogisticssystemincludedozensofdifferentforecastingalgorithmsthattheanalystcanusetogeneratealternativedemandforecasts.Whilescoresofdifferentforecastingtechniquesexist,almostanyforecastingprocedurecanbebroadlyclassifiedintooneofthefollowingfourbasiccategoriesbasedonthefundamentalapproachtowardstheforecastingproblemthatisemployedbythetechnique.JudgmentalApproaches.Theessenceofthejudgmentalapproachistoaddresstheforecastingissuebyassumingthatsomeoneelseknowsandcantellyoutherightanswer.Thatis,inajudgment-basedtechniquewegathertheknowledgeandopinionsofpeoplewhoareinapositiontoknowwhatdemandwillbe.Forexample,wemightconductasurveyofthecustomerbasetoestimatewhatoursaleswillbenextmonth.Bytheirnature,judgment-basedforecastsusesubjectiveandqualitativedatatoforecastfutureoutcomes.Theyinherentlyrelyonexpertopinion,experience,judgment,intuition,conjecture,andother"soft"data.Suchtechniquesareoftenusedwhenhistoricaldataarenotavailable,asisthecasewiththeintroductionofanewproductorservice,andinforecastingtheimpactoffundamentalchangessuchasnewtechnologies,environmentalchanges,culturalchanges,legalchanges,andsoforth.Someofthemorecommonproceduresincludethefollowing:1.Surveys.Thisisa"bottomup"approachwhereeachindividualcontributesapieceofwhatwillbecomethefinalforecast.Forexample,wemightpollorsampleourcustomerbasetoestimatedemandforacomingperiod.Alternatively,wemightgatherestimatesfromoursalesforceastohowmucheachsalespersonexpectstosellinthenexttimeperiod.Theapproachisatleastplausibleinthesensethatweareaskingpeoplewhoareinapositiontoknowsomethingaboutfuturedemand.Ontheotherhand,inpracticetherehaveproventobeseriousproblemsofbiasassociatedwiththesetools.Itcanbedifficultandexpensivetogatherdatafromcustomers.Historyalsoshowsthatsurveysof"intentiontopurchase"willgenerallyover-estimateactualdemand-ikingaproductisonething,butactuallybuyingitisoftenquiteanother.Salespeoplemayalsointentionally(orevenunintentionally)exaggerateorunderestimatetheirsalesforecastsbasedonwhattheybelievetheirsupervisorswantthemtosay.Ifthesalesforce(orthecustomerbase)believesthattheirforecastswilldeterminetheleveloffinishedgoodsinventorythatwillbeavailableinthenextperiod,theymaybesorelytemptedtoinflatetheirdemandestimatessoastoinsuregoodinventoryavailability.Evenifthesebiasescouldbeeliminatedorcontrolled,anotherseriousproblemwouldprobablyremain.Salespeoplemightbeabletoestimatetheirweeklydollarvolumeortotalunitsales,buttheyarenotlikelytobeabletodevelopcredibleestimatesattheSKUlevelthatthelogisticssystemwillrequire.Forthesereasonsitwillseldombethecasethatthesetoolswillformthebasisofasuccessfuldemandforecastingprocedureinalogisticssystem.Consensusmethods.Asanalternativetothe"bottom-up"surveyapproaches,consensusmethodsuseasmallgroupofindividualstodevelopgeneralforecasts.Ina“JuryofExecutiveOpinion”,forexample,agroupofexecutivesinthefirmwouldmeetanddevelopthroughdebateanddiscussionageneralforecastofdemand.Eachindividualwouldpresumablycontributeinsightandunderstandingbasedontheirviewofthemarket,theproduct,thecompetition,andsoforth.Onceagain,whiletheseexecutivesareundoubtedlyexperienced,theyarehardlydisinterestedobservers,andtheopportunityforbiasedinputsisobvious.Amoreformalconsensusprocedure,called“TheDelphiMethod”,hasbeendevelopedtohelpcontroltheseproblems.Inthistechnique,apanelofdisinterestedtechnicalexpertsispresentedwithaquestionnaireregardingaforecast.Theanswersarecollected,processed,andre-distributedtothepanel,makingsurethatallinformationcontributedbyanypanelmemberisavailabletoallmembers,butonananonymousbasis.Eachexpertreflectsonthegatheringopinion.Asecondquestionnaireisthendistributedtothepanel,andtheprocessisrepeateduntilaconsensusforecastisreached.Consensusmethodsareusuallyappropriateonlyforhighlyaggregateandusuallyquitelong-rangeforecasts.Onceagain,theirabilitytogenerateusefulSKUlevelforecastsisquestionable,anditisunlikelythatthisapproachwillbethebasisforasuccessfuldemandforecastingprocedureinalogisticssystem.Judgment-basedmethodsareimportantinthattheyareoftenusedtodetermineanenterprise'sstrategy.Theyarealsousedinmoremundanedecisions,suchasdeterminingthequalityofapotentialvendorbyaskingforreferences,andtherearemanyotherreasonableapplications.Itistruethatjudgmentbasedtechniquesareaninadequatebasisforademandforecastingsystem,butthisshouldnotbeconstruedtomeanthatjudgmenthasnoroletoplayinlogisticsforecastingorthatsalespeoplehavenoknowledgetobringtotheproblem.Infact,itisoftenthecasethatsalesandmarketingpeoplehavevaluableinformationaboutsalespromotions,newproducts,competitoractivity,andsoforth,whichshouldbeincorporatedintotheforecastsomehow.Manyorganizationstreatsuchdataasadditionalinformationthatisusedtomodifytheexistingforecastratherthanasthebaselinedatausedtocreatetheforecastinthefirstplace.ExperimentalApproaches.Anotherapproachtodemandforecasting,whichisappealingwhenanitemis"new"andwhenthereisnootherinformationuponwhichtobaseaforecast,istoconductademandexperimentonasmallgroupofcustomersandtoextrapolatetheresultstoalargerpopulation.Forexample,firmswilloftentestanewconsumerproductinageographicallyisolated"testmarket"toestablishitsprobablemarketshare.Thisexperienceisthenextrapolatedtothenationalmarkettoplanthenewproductlaunch.Experimentalapproachesareveryusefulandnecessaryfornewproducts.Butforexistingproductsthathaveanaccumulatedhistoricaldemandrecorditseemsintuitivethatdemandforecastsshouldsomehowbebasedonthisdemandexperience.Formostfirms(withsomeverynotableexceptions)thelargemajorityofSKUsintheproductlinehavelongdemandhistories.Intheearlystagesofnewproductdevelopmentitisimportanttogetsomeestimateofthelevelofpotentialdemandfortheproduct.Avarietyofmarketresearchtechniquesareusedtothisend.CustomerSurveysaresometimesconductedoverthetelephoneoronstreetcorners,atshoppingmalls,andsoforth.Thenewproductisdisplayedordescribed,andpotentialcustomersareaskedwhethertheywouldbeinterestedinpurchasingtheitem.Whilethisapproachcanhelptoisolateattractiveorunattractiveproductfeatures,experiencehasshownthat"intenttopurchase"asmeasuredinthiswayisdifficulttotranslateintoameaningfuldemandforecast.Thisfallsshortofbeingatrue“demandexperiment”.ConsumerPanelsarealsousedintheearlyphasesofproductdevelopment.Hereasmallgroupofpotentialcustomersarebroughttogetherinaroomwheretheycanusetheproductanddiscussitamongthemselves.Panelmembersareoftenpaidanominalamountfortheirparticipation.Likesurveys,theseproceduresaremoreusefulforanalyzingproductattributesthanforestimatingdemand,andtheydonotconstitutetrue“demandexperiments”becausenopurchasestakeplace.TestMarketingisoftenemployedafternewproductdevelopmentbutpriortoafull-scalenationallaunchofanewbrandorproduct.Theideaistochoosearelativelysmall,reasonablyisolated,yetsomehowdemographically"typical"marketarea.IntheUnitedStates,thisisoftenamediumsizedcitysuchasCincinnatiorBuffalo.Thetotalmarketingplanfortheitem,includingadvertising,promotions,anddistributiontactics,is"rolledout"andimplementedinthetestmarket,andmeasurementsofproductawareness,marketpenetration,andmarketsharearemade.Whilethesedataareusedtoestimatepotentialsalestoalargernationalmarket,theemphasishereisusuallyon"fine-tuning"thetotalmarketingplanandinsuringthatnoproblemsorpotentialembarrassmentshavebeenoverlooked.Forexample,andextensivelytest-marketeditsPringlespotatochipproductmadewiththefatsubstituteOlestratoassurethattheproductwouldbebroadlyacceptabletothemarket.Relational/CausalApproaches.Theassumptionbehindacausalorrelationalforecastisthat,simplyput,thereisareasonwhypeoplebuyourproduct.Ifwecanunderstandwhatthatreason(orsetofreasons)is,wecanusethatunderstandingtodevelopademandforecast.Forexample,ifwesellumbrellasatasidewalkstand,wewouldprobablynoticethatdailydemandisstronglycorrelatedtotheweather-wesellmoreumbrellaswhenitrains.Oncewehaveestablishedthisrelationship,agoodweatherforecastwillhelpusorderenoughumbrellastomeettheexpecteddemand."TimeSeries"Approaches.Atimeseriesprocedureisfundamentallydifferentthanthefirstthreeapproacheswehavediscussed.Inapuretimeseriestechnique,nojudgmentorexpertiseoropinionissought.Wedonotlookfor"causes"orrelationshipsorfactorswhichsomehow"drive"demand.Wedonottestitemsorexperimentwithcustomers.Bytheirnature,timeseriesproceduresareappliedtodemanddatathatarelongitudinalratherthancross-sectional.Thatis,thedemanddatarepresentexperiencethatisrepeatedovertimeratherthanacrossitemsorlocations.Theessenceoftheapproachistorecognize(orassume)thatdemandoccursovertimeinpatternsthatrepeatthemselves,atleastapproximately.Ifwecandescribethesegeneralpatternsortendencies,withoutregardtotheir"causes",wecanusethisdescriptiontoformthebasisofaforecast.Inonesense,allforecastingproceduresinvolvetheanalysisofhistoricalexperienceintopatternsandtheprojectionofthosepatternsintothefutureinthebeliefthatthefuturewillsomehowresemblethepast.Thedifferencesinthefourapproachesareinthewaythis"searchforpattern"isconducted.Judgmentalapproachesrelyonthesubjective,ad-hocanalysesofexternalindividuals.Experimentaltoolsextrapolateresultsfromsmallnumbersofcustomerstolargepopulations.Causalmethodssearchforreasonsfordemand.Timeseriestechniquessimplyanalyzethedemanddatathemselvestoidentifytemporalpatternsthatemergeandpersist.外文翻譯1:需求預(yù)測需求預(yù)測的重要性產(chǎn)品需求預(yù)測,對任何供應(yīng)商、制造商或零售商都是至關(guān)重要的。預(yù)測未來的需求將決定應(yīng)購買、生產(chǎn)、出貨的數(shù)量。需求預(yù)測是必要的,因為基本操作過程中,從供應(yīng)商的原材料到成品的商品,到顧客手中,需要一定的時間。大多數(shù)的企業(yè)不能簡單地等待需求出現(xiàn),然后再作出反應(yīng)。相反,他們必須預(yù)先考慮并規(guī)劃未來的需求,使他們能在客戶訂單出現(xiàn)后立即作出反應(yīng)?;蛘邠Q句話說,大多數(shù)廠家是“為庫存”,而不是“為訂單”而生產(chǎn)——他們提前計劃,然后再調(diào)配庫存產(chǎn)成品到目標(biāo)地點(diǎn)。因此,一旦客戶訂單出現(xiàn),它就可以完成。由于大多數(shù)用戶不愿意花時間等待,在實(shí)際過程中,他們希望整個供應(yīng)鏈根據(jù)他們的訂單來運(yùn)轉(zhuǎn)。一個訂單周期可能需要數(shù)周或數(shù)月,從供應(yīng)商和分包裝配,到制造產(chǎn)品,再經(jīng)過最終裝運(yùn)的訂單到達(dá)客戶。公司提供快速交貨到客戶往往會迫使所有的競爭者在市場上保持充足的成品庫存,以提供快速訂單周期。因此,幾乎每一個組織包括制造部門或訂單的基礎(chǔ)是對未來的需求預(yù)測。能夠準(zhǔn)確預(yù)測需求,給公司機(jī)會通過確定水準(zhǔn)測量與生產(chǎn)數(shù)量、理順交通運(yùn)輸、規(guī)劃高效率的物流運(yùn)作來控制成本。一般的做法,準(zhǔn)確的需求預(yù)測能導(dǎo)致有效率的運(yùn)作和高水平的客戶服務(wù),同時不準(zhǔn)確的預(yù)測,將不可避免地導(dǎo)致效率低下成本高的運(yùn)作和低水準(zhǔn)服務(wù)。在許多供應(yīng)鏈中,我們提高物流過程效率和效能的關(guān)鍵是要提高對需求的預(yù)測水平。預(yù)測的一般方法所有公司都預(yù)測需求,但很難找到兩家用一種需求預(yù)測方法的公司。在過去的幾十年中,有很多不同的預(yù)測方法運(yùn)用于若干不同的應(yīng)用領(lǐng)域,包括工程和經(jīng)濟(jì)學(xué)。很多這樣的方法已經(jīng)應(yīng)用到實(shí)際的問題中,對需求的預(yù)測在物流系統(tǒng)中,取得了不同程度的成功。大多數(shù)商業(yè)軟件包支持需求預(yù)測,在物流系統(tǒng)中包括幾十種不同的預(yù)測算法,分析師可以使用產(chǎn)生替代性需求的預(yù)測。雖然數(shù)十種不同的預(yù)測技術(shù)存在,但幾乎任何一個預(yù)測的程序大致可分到下列四項基本分類:1、主觀判斷的辦法。這種判斷方法的本質(zhì)是,假設(shè)有其它人知道,并能告訴你正確的答案。也就是通過以判斷為基礎(chǔ)的技術(shù),我們收集了解需求的人的見解。舉例來說,我們可能會進(jìn)行一項調(diào)查,以客戶基礎(chǔ),估計我們在下個月的銷售收入。就其性質(zhì)而言,是判斷為基礎(chǔ)的預(yù)測,用主觀和定性數(shù)據(jù),以預(yù)測未來的結(jié)果。他們本身就依靠專家的意見、經(jīng)驗、判斷、直覺、猜想等“軟”數(shù)據(jù)。這些技巧經(jīng)常被用在沒有歷史數(shù)據(jù)情況下,由于引入了新的產(chǎn)品或服務(wù),對預(yù)測的影響發(fā)生了根本變化,如新的技術(shù)、環(huán)境的改變、文化的變化、法律的變化等。一些比較常用的程序包括以下幾個方面:(1)、調(diào)查。這是一種“自下而上”的辦法,每一個人的貢獻(xiàn)一部分,將成為最后的預(yù)測。舉例來說,我們可能會以我們的客戶基礎(chǔ)為樣本,來估計今后一個時期的需求?;蛘?,我們可能會收集我們的銷售隊伍每個營業(yè)員今后一段時期預(yù)計的銷售。這種做法至少在某種意義上說,我們從處在那個位置的人了解一些有關(guān)未來的需求。在另一方面,在實(shí)踐中有已經(jīng)被證明是存在嚴(yán)重的偏見的這些工具可能導(dǎo)致困難和昂貴的花銷從顧客收集數(shù)據(jù)。歷史還表明,調(diào)查“打算購買”通常會高估實(shí)際需求-喜歡一個產(chǎn)品是一回事,但實(shí)際上購買它往往是另一回事。銷售人員也可能故意(甚至有意無意)夸大或低估自己的銷售預(yù)測,他們相信他們的主管希望聽到他們說的。如果銷售人員(或客戶群)認(rèn)為,他們的預(yù)測將決定今后一個時期成品庫存的水平,他們可能被誘惑提高預(yù)期需求和預(yù)算,以確保良好存貨可用性。即使這些偏見是可以消除或控制的,另一個嚴(yán)重問題,將很可能維持不變。銷售人也許能估計其每周銷售量或總銷售量,但是卻不太可能在存貨單元一級發(fā)展對物流系統(tǒng)可信的預(yù)測?;谶@些原因,如果情況屬實(shí),會很難在這些工具的基礎(chǔ)上形成一個對物流系統(tǒng)成功的需求預(yù)測程序。(2)、協(xié)商的方法。作為一種替代,以“自下而上”的統(tǒng)計調(diào)查、協(xié)商的方法,利用一小群個人預(yù)測來發(fā)展一般的預(yù)測。在“陪審團(tuán)的執(zhí)行意見”里,舉例來說,某集團(tuán)總裁在該公司將通過辯論和討論一般預(yù)測的需求來滿足和發(fā)展。在洞察和理解的基礎(chǔ)上,每一個人的看法、市場、產(chǎn)品、競爭觀念等等大概將有助于他們。再次,盡管這些高管,無疑是有經(jīng)驗的,但他們不是超然的觀察員,并顯而易見的有機(jī)會錯誤的投入。一個較正式的協(xié)商程序,是以所謂的“德爾菲法”幫助控制這些問題。在這種技術(shù)的一個小組,向超然的技術(shù)專家介紹同一份問卷以進(jìn)行預(yù)測。答案經(jīng)過采集、處理,并重新分發(fā)到小組,確保提供給所有成員,但需在一個匿名的基礎(chǔ)上。每個專家對所收集的意見進(jìn)行反映。第二次問卷調(diào)查再分發(fā)給各小組,反復(fù)這個過程,直至達(dá)成預(yù)測共識。達(dá)成共識的方法,通常只適宜高度總結(jié)并需經(jīng)過相當(dāng)長的時間。再次,他們的能力差別使在存貨單元水平對物流系統(tǒng)的需求預(yù)測難度很大。判斷為基礎(chǔ)的方法是重要的,因為它們常常用來決定一個企業(yè)的戰(zhàn)略。它們也可用于更為簡單平凡的決策,如確定一個潛在的供應(yīng)商所要求的質(zhì)量的參考作用,而且還有許多其他的合理應(yīng)用。以判斷為基礎(chǔ)的技術(shù)對基礎(chǔ)需求預(yù)測系統(tǒng)是不夠的,但是這不應(yīng)當(dāng)被理解為,判斷在物流預(yù)測或銷售人員中并沒有發(fā)揮作用。事實(shí)上,在很多情況下,銷售和營銷人的寶貴資料、促銷活動、新產(chǎn)品、競爭對手的活動等等應(yīng)或多或少納入預(yù)測。許多組織把這些數(shù)據(jù)作為補(bǔ)充資料,是用來修改現(xiàn)有的預(yù)測,而不是作為基準(zhǔn)數(shù)據(jù)來建立預(yù)測。2、實(shí)驗方法。需求預(yù)測的另一種做法是,當(dāng)一個項目是“新”的時,沒有其他資料為根據(jù)進(jìn)行預(yù)測,只能對一小群客戶進(jìn)行實(shí)驗,并根據(jù)推斷結(jié)果推廣。例如,公司經(jīng)常會為考察一個新的消費(fèi)產(chǎn)品,在地理上孤立出“試驗市場”,以確定其可能的市場份額。這一經(jīng)驗,然后推廣到全國市場計劃,進(jìn)行新產(chǎn)品的推出。實(shí)驗的做法是十分有益和必要的,但對有需求紀(jì)錄的現(xiàn)有產(chǎn)品,似乎直觀的需求預(yù)測應(yīng)該或多或少在需求的經(jīng)驗基礎(chǔ)上。對于大多數(shù)企業(yè)(一些十分顯著的例外),大部分的產(chǎn)品,在產(chǎn)品線上有長期的需求的歷史。在新產(chǎn)品開發(fā)的早期階段,得到一些人估計的產(chǎn)品的潛在需求水平是非常重要的。不同的市場研究技術(shù)用于這一目的。在顧客意見調(diào)查中,有時進(jìn)行電話調(diào)查或在商場等地方進(jìn)行街頭問卷調(diào)查。對潛在的客戶進(jìn)行新產(chǎn)品展示或描述,并問他們是否有興趣購買該項目。這種做法雖然可以幫助區(qū)分不同產(chǎn)品間吸引力的不同,但經(jīng)驗表明,以“采購意向”為衡量標(biāo)準(zhǔn),很難轉(zhuǎn)化為有意義的需求預(yù)測。這屬于短期的是一個真正的“需求”實(shí)驗。消費(fèi)問題小組還用在早期階段的產(chǎn)品開發(fā)工作。一小群潛在的客戶聚集在一個房間里,他們可以使用該產(chǎn)品,并討論它們。小組成員往往是為他們的參與付出了面值。像調(diào)查中,這些過程都有益于分析產(chǎn)品屬性以估算需求,但們不構(gòu)成真正的“需求實(shí)驗”,因為沒有購買的發(fā)生。測試市場往往是在新產(chǎn)品開發(fā)后使用,但事先必須在全國大規(guī)模推出新的品牌或產(chǎn)品。我們的構(gòu)想是選擇一個相對較小,合理孤立的“典型”的市場領(lǐng)域。在美國,這往往是一個中型城市如辛辛那提或水牛??偟臓I銷計劃,包括廣告、促銷及分銷策略,是“滾出來”并落實(shí)在測試市場,以及度量取得的產(chǎn)品知名度,市場滲透和市場占有率均。這些數(shù)據(jù)被用來估計更大的國內(nèi)市場潛在銷售,重點(diǎn)通常是“微調(diào)”整體市場營銷計劃,并確保沒有任何潛在問題被忽略。舉例來說,Proctor和Gamble廣泛推銷它的Pringles土豆片生產(chǎn)與脂肪替代品olestra,以保證該產(chǎn)品將被市場廣泛接受。3、關(guān)聯(lián)/因果法。在一個原因后的假設(shè),簡單地說,別人買我們的產(chǎn)品是有原因的。如果我們能夠理解這個原因(或一套理由),我們可以利用它來制定一項需求預(yù)測。舉例來說,如果我們賣雨傘,在行人的立場上,我們可能會看到每天的需求是與天氣密切相關(guān)的-我們在下雨時銷售更多的雨傘。一旦我們建立了這種關(guān)系,一個良好的氣象預(yù)報將有助于我們準(zhǔn)備足夠的雨傘,以應(yīng)付預(yù)期的需求。4、“時間序列”的方法。時間序列的方法,和第一、三種我們已討論的方法從根本上不同。一個純粹的時間序列技術(shù),沒有任何判斷或意見。我們不找在某種程度上“驅(qū)動”需求的“原因”或關(guān)系的因素,我們不檢驗項目或?qū)嶒炁c客戶的聯(lián)系。就其性質(zhì)而言,時間序列的程序,適用于需求數(shù)據(jù)是縱向而非橫截面。這就是說,需求數(shù)據(jù)所代表的經(jīng)驗,是一再重復(fù)的,而不是采取一刀切的物品或地點(diǎn)。本質(zhì)的做法,是承認(rèn)(或假設(shè))需求發(fā)生的時間在重復(fù)的模式。如果我們能夠描述這些一般模式或傾向,而不考慮其“原因”,我們可以在此基礎(chǔ)上形成的一個預(yù)測。在某種意義上說,所有的預(yù)測程序涉及對歷史的經(jīng)驗的分析,并預(yù)測未來,我們的前景有點(diǎn)類似于過去。這四種辦法的不同,是“尋找”的模式不同。主觀判斷的辦法,依賴于主觀的,特設(shè)分析外部的個人。實(shí)驗工具推斷結(jié)果,從少量的顧客推廣到大批人群。因果方法尋找需求原因。時間序列技術(shù)淺析需求數(shù)據(jù)本身,以確定時序模式出現(xiàn),并持續(xù)下去。英語原文2:TheABCsofSupplyChainManagement

ByChristopherKochWhatissupplychainmanagement?Supplychainmanagementisthecombinationofartandsciencethatgoesintoimprovingthewayyourcompanyfindstherawcomponentsitneedstomakeaproductorservice,manufacturesthatproductorserviceanddeliversittocustomers.Thefollowingarefivebasiccomponentsforsupplychainmanagement.1.Plan-Thisisthestrategicportionofsupplychainmanagement.Youneedastrategyformanagingalltheresourcesthatgotowardmeetingcustomerdemandforyourproductorservice.Abigpieceofplanningisdevelopingasetofmetricstomonitorthesupplychainsothatitisefficient,costslessanddelivershighqualityandvaluetocustomers.2.Source-Choosethesuppliersthatwilldeliverthegoodsandservicesyouneedtocreateyourproductorservice.Developasetofpricing,deliveryandpaymentprocesseswithsuppliersandcreatemetricsformonitoringandimprovingtherelationships.Andputtogetherprocessesformanagingtheinventoryofgoodsandservicesyoureceivefromsuppliers,includingreceivingshipments,verifyingthem,transferringthemtoyourmanufacturingfacilitiesandauthorizingsupplierpayments.Make-Thisisthemanufacturingstep.Scheduletheactivitiesnecessaryforproduction,testing,packagingandpreparationfordelivery.Asthemostmetric-intensiveportionofthesupplychain,measurequalitylevels,productionoutputandworkerproductivity.Deliver-Thisisthepartthatmanyinsidersrefertoas"logistics."Coordinatethereceiptofordersfromcustomers,developanetworkofwarehouses,pickcarrierstogetproductstocustomersandsetupaninvoicingsystemtoreceivepayments.Return-Theproblempartofthesupplychain.Createanetworkforreceivingdefectiveandexcessproductsbackfromcustomersandsupportingcustomerswhohaveproblemswithdeliveredproducts.Whatdoessupplychainmanagementsoftwaredo?Supplychainmanagementsoftwareispossiblythemostfracturedgroupofsoftwareapplicationsontheplanet.Eachofthefivemajorsupplychainstepspreviouslyoutlinedcomposesdozensofspecifictasks,manyofwhichhavetheirownspecificsoftware.Therearesomelargevendorsthathaveattemptedtoassemblemanyofthesedifferentchunksofsoftwaretogetherunderasingleroof,butnoonehasacompletepackage.Integratingthedifferentsoftwarepiecestogethercanbeanightmare.Perhapsthebestwaytothinkaboutsupplychainsoftwareistoseparateitintosoftwarethathelpsyouplanthesupplychainandsoftwarethathelpsyouexecutethesupplychainstepsthemselves.Supplychainplanning(SCP)softwareusesfancymathalgorithmstohelpyouimprovetheflowandefficiencyofthesupplychainandreduceinventory.SCPisentirelydependentuponinformationforitsaccuracy.Ifyou'reamanufacturerofconsumerpackagedgoodsforexample,don'texpectyourplanningapplicationstobeveryaccurateifyoucan'tfeedthemaccurate,up-to-dateinformationaboutcustomerordersfromyourretailcustomers,salesdatafromyourretailercustomers'stores,manufacturingcapacityanddeliverycapability.Thereareplanningapplicationsavailableforallfiveofthemajorsupplychainstepspreviouslylisted.Arguablythemostvaluable(andcomplexandpronetoerror)isdemandplanning,whichdetermineshowmuchproductyouwillmaketosatisfyyourdifferentcustomers'demands.Supplychainexecution(SCE)softwareisintendedtoautomatethedifferentstepsofthesupplychain.Thiscouldbeassimpleaselectronicallyroutingordersfromyourmanufacturingplantstoyoursuppliersforthestuffyouneedtomakeyourproducts.DoIneedtohaveERPsoftwarebeforeIinstallsupplychainsoftware?Thisisaverycontroversialsubject.YoumayneedERPifyouplantoinstallSCPapplicationsbecausetheyarereliantuponthekindofinformationthatisstoredinthemostquantityinsideERPsoftware.TheoreticallyyoucouldassembletheinformationyouneedtofeedtheSCPapplicationsfromlegacysystems(formostcompaniesthismeansExcelspreadsheetsspreadoutallovertheplace),butitcanbenightmarishtotrytogetthatinformationflowingonafast,reliablebasisfromalltheareasofthecompany.ERPisthebatteringramthatintegratesallthatinformationtogetherinasingleapplication,andSCPapplicationsbenefitfromhavingasinglemajorsourcetogotoforup-to-dateinformation.MostCIOswhohavetriedtoinstallSCPapplicationssaytheyaregladtheydidERPfirst.TheycalltheERPprojects"puttingyourinformationhouseinorder."Ofcourse,ERPisexpensiveanddifficult,soyoumaywanttoexplorewaystofeedyourSCPapplicationstheinformationtheyneedwithoutdoingERPfirst.SCEapplicationsarelessdependentupongatheringinformationfromaroundthecompany,sotheytendtobeindependentoftheERPdecision.Butchancesare,you'llneedtohavetheSCEapplicationscommunicatewithERPinsomefashion.It'simportanttopayattentiontoSCEsoftware'sabilitytointegratewiththeInternetandwithERPorSCPapplicationsbecausetheInternetwilldrivedemandforintegratedinformation.Forexample,ifyouwanttobuildaprivatewebsiteforcommunicatingwithyourcustomersandsuppliers,youwillwanttopullinformationfromSCE,SCPandERPapplicationstogethertopresentupdatedinformationaboutorders,payments,manufacturingstatusanddelivery.Whatisthegoalofinstallingsupplychainmanagementsoftware?BeforetheInternetcamealong,theaspirationsofsupplychainsoftwaredevoteeswerelimitedtoimprovingtheirabilitytopredictdemandfromcustomersandmaketheirownsupplychainsrunmoresmoothly.Butthecheap,ubiquitousnatureoftheInternet,alongwithitssimple,universallyacceptedcommunicationstandardshavethrownthingswideopen.Now,theoreticallyanyway,youcanconnectyoursupplychainwiththesupplychainsofyoursuppliersandcustomerstogetherinasinglevastnetworkthatoptimizescostsandopportunitiesforeveryoneinvolved.ThiswasthereasonfortheB2Bexplosion;theideathateveryoneyoudobusinesswithcouldbeconnectedtogetherintoonebighappy,cooperativefamily.Ofcourse,therealitybehindthisvisionisthatitwilltakeyearstocometofruition.ButconsideringthatB2Bhasonlybeenaroundforafewyears,someindustrieshavealreadymadegreatprogress,mostnotablyconsumer-packagedgoods(thecompaniesthatmakeproductsthatgotosupermarketsanddrugstores),hightechnologyandautos.Whenyouaskthepeopleonthefrontlinesintheseindustrieswhattheyhopetogainfromtheirsupplychaineffortsinthenearterm,theywillallrespondwithasingleword:visibility.Thesupplychaininmostindustriesislikeabigcardgame.Theplayersdon'twanttoshowtheircardsbecausetheydon'ttrustanyoneelsewiththeinformation.Butiftheyshowedtheirhandstheycouldallbenefit.Supplierswouldn'thavetoguesshowmuchrawmaterialstoorder,andmanufacturerswouldn'thavetoordermorethantheyneedfromsupplierstomakesuretheyhaveenoughonhandifdemandfortheirproductsunexpectedlygoesup.Andretailerswouldhavefeweremptyshelvesiftheysharedtheinformationtheyhadaboutsalesofamanufacturer'sproductinalltheirstoreswiththemanufacturer.TheInternetmakesshowingyourhandtootherspossible,butcenturiesofdistrustandlackofcoordinationwithinindustriesmakeitdifficult.Whatissupplychaincollaboration?Let'slookatconsumerpackagedgoodsasanexampleofcollaboration.Iftherearetwocompaniesthathavemadesupplychainahouseholdword,theyareWal-MartandProcter&Gamble.Beforethesetwocompaniesstartedcollaboratingbackinthe'80s,retailerssharedverylittleinformationwithmanufacturers.ButthenthetwogiantsbuiltasoftwaresystemthathookedP&GuptoWal-Mart'sdistributioncenters.WhenP&G'sproductsrunlowatthedistributioncenters,thesystemsendsanautomaticalerttoP&Gtoshipmoreproducts.Insomecases,thesystemgoesallthewaytotheindividualWal-Martstore.ItletsP&Gmonitortheshelvesthroughreal-timesatellitelink-upsthatsendmessagestothefactorywheneveraP&Gitemswoopspastascannerattheregister.Withthiskindofminute-to-minuteinformation,P&Gknowswhentomake,shipanddisplaymoreproductsattheWal-Martstores.NoneedtokeepproductspiledupinwarehousesawaitingWal-Mart'scall.Invoicingandpaymentshappenautomaticallytoo.ThesystemsavesP&Gsomuchintime,reducedinventoryandlowerorder-processingcoststhatitcanaffordtogiveWal-Mart"low,everydayprices"withoutputtingitselfoutofbusiness.CiscoSystems,whichmakesequipmenttohookuptotheInternet,isalsofamousforitssupplychaincollaboration.Ciscohasanetworkofcomponentsuppliers,distributorsandcontractmanufacturersthatarelinkedthroughCisco'sextranettoformavirtual,just-in-timesupplychain.WhenacustomerordersatypicalCiscoproduct-forexample,arouterthatdirectsInternettrafficoveracompanynetwork-throughCisco'swebsite,theordertriggersaflurryofmessagestocontractmanufacturersofprintedcircuitboardassemblies.Distributors,meanwhile,arealertedtosupplythegenericcomponentsoftherouter,suchasapowersupply.Cisco'scontractmanufacturers,someofwhommakesubassembliesliketherouterchassisandotherswhoassemblethefinishedproduct,alreadyknowwhat'scomingdowntheorderpipebecausethey'veloggedontoCisco'sextranetandlinkedintoCisco'sownmanufacturingexecutionsystems.SoonafterthecontractmanufacturersreachintoCisco'sextranet,theextranetstartspokingaroundthecontractor'sassemblylinetomakesureeverythingis

溫馨提示

  • 1. 本站所有資源如無特殊說明,都需要本地電腦安裝OFFICE2007和PDF閱讀器。圖紙軟件為CAD,CAXA,PROE,UG,SolidWorks等.壓縮文件請下載最新的WinRAR軟件解壓。
  • 2. 本站的文檔不包含任何第三方提供的附件圖紙等,如果需要附件,請聯(lián)系上傳者。文件的所有權(quán)益歸上傳用戶所有。
  • 3. 本站RAR壓縮包中若帶圖紙,網(wǎng)頁內(nèi)容里面會有圖紙預(yù)覽,若沒有圖紙預(yù)覽就沒有圖紙。
  • 4. 未經(jīng)權(quán)益所有人同意不得將文件中的內(nèi)容挪作商業(yè)或盈利用途。
  • 5. 人人文庫網(wǎng)僅提供信息存儲空間,僅對用戶上傳內(nèi)容的表現(xiàn)方式做保護(hù)處理,對用戶上傳分享的文檔內(nèi)容本身不做任何修改或編輯,并不能對任何下載內(nèi)容負(fù)責(zé)。
  • 6. 下載文件中如有侵權(quán)或不適當(dāng)內(nèi)容,請與我們聯(lián)系,我們立即糾正。
  • 7. 本站不保證下載資源的準(zhǔn)確性、安全性和完整性, 同時也不承擔(dān)用戶因使用這些下載資源對自己和他人造成任何形式的傷害或損失。

評論

0/150

提交評論