泵體加工專用機(jī)床設(shè)計(jì)—鉆8-M10底孔組合機(jī)床設(shè)計(jì)
資源目錄里展示的全都有，所見即所得。下載后全都有,請(qǐng)放心下載。原稿可自行編輯修改=【QQ：401339828 或11970985 有疑問可加】
附錄1 外文翻譯
基于事例推理的夾具設(shè)計(jì)研究與應(yīng)用
摘要:根據(jù)基于事例的設(shè)計(jì)方法,提出采用工序件的特征信息和夾具的結(jié)構(gòu)特征信息來(lái)描述夾具的相似性,并建立了包括這2方面主要特征信息為基礎(chǔ)的事例索引碼,設(shè)計(jì)了事例庫(kù)的結(jié)構(gòu)形式,創(chuàng)建了層次化的事例組織方式;同時(shí),提出了基于知識(shí)引導(dǎo)的夾具事例檢索算法,以及事例的修改和采用同族事例碼進(jìn)行相似事例的存貯,形成了基于事例推理的夾具設(shè)計(jì).所開發(fā)的原型系統(tǒng)在型號(hào)工程夾具設(shè)計(jì)等項(xiàng)目的設(shè)計(jì)過程中得到了應(yīng)用,并取得了令人滿意的使用效果.
關(guān)鍵詞: 基于事例的推理 夾具設(shè)計(jì) CAD
夾具是以確定工件安全定位準(zhǔn)確為目的的裝置，并在加工過程中保持工件與刀具或機(jī)床的位置一致不變。因?yàn)閵A具的結(jié)構(gòu)依賴于產(chǎn)品的特點(diǎn)和在企業(yè)規(guī)劃中加工工序的地位,所以它的設(shè)計(jì)是制造過程中的瓶頸,制約著效率的提高. 夾具設(shè)計(jì)是一個(gè)復(fù)雜的過程,需要有從大量的設(shè)計(jì)論文中了解質(zhì)量知識(shí)的經(jīng)驗(yàn)，這些設(shè)計(jì)論文包括工件的結(jié)構(gòu)設(shè)計(jì)、涉及加工工藝，和加工環(huán)境。當(dāng)用這些擅長(zhǎng)繪制詳細(xì)設(shè)計(jì)圖的傳統(tǒng)的CAD工具（如Unigraphics、CATIA、Pro/E)時(shí),這仍然是一項(xiàng)非常耗時(shí)的工作，但是利用以往的設(shè)計(jì)經(jīng)驗(yàn)和資源也不能提供一些益處,而這正是提高效率的關(guān)鍵因素. 基于事例推理 (CBR) 的方法適應(yīng)以往個(gè)案解決的辦法,建立一個(gè)新問題的方法,主要有以下四步驟:檢索、利用、修改,并保留.這是一個(gè)比用專業(yè)系統(tǒng)模仿人類思維有用的使用方法，因?yàn)樘岢鲆粋€(gè)類似的情況,和采用一些修改,似乎不言自明,而且比人類更直觀.所以支持不同事例的設(shè)計(jì)工具已經(jīng)在諸多領(lǐng)域中發(fā)展起來(lái)，如在注射成型及設(shè)計(jì)、建筑設(shè)計(jì)、模具設(shè)計(jì)投死, 規(guī)劃過程中,還有夾具設(shè)計(jì). 孫用六個(gè)數(shù)字組成代碼參數(shù),包括工件的形狀、機(jī)械部分、軸襯,第一定位裝置,第二定位裝置和夾緊裝置. 但這個(gè)系統(tǒng)不能用于除鉆床夾具外的其他夾具類型,不能解決儲(chǔ)存需要保留的同一參數(shù)代碼的問題,這在CBR中是非常重要的.
1事例參數(shù)和事例圖書館的建立
1.1事例參數(shù)
事例參數(shù)應(yīng)該由工件的所有的特征組成，來(lái)區(qū)別不同的夾具. 使用他們能夠使操作方便. 因?yàn)榱慵男螤钍嵌喾N多樣的, 在生產(chǎn)企業(yè)中制造的技術(shù)要求也不斷發(fā)展,許多特征作被用做事例參數(shù)將會(huì)使搜索速度降低,其主要特征是不重要的,因?yàn)榉峙浣o每個(gè)特征的比重必須減少. 另一方面,事例參數(shù)包含所有的特征是困難的。
因此,考慮到實(shí)際和快速設(shè)計(jì)的需求,事例參數(shù)要包含工件的主要特征和夾具的結(jié)構(gòu)。事例參數(shù)代碼由16位數(shù)組成：13位數(shù)是事例特征 3位數(shù)是事例識(shí)別數(shù)字。
前13位數(shù)代表13個(gè)特征。 每個(gè)數(shù)字與特征的一個(gè)屬性相一致,這可能是"*"、"？"、"1"、"2"，…，"A"、"B",…，"Z"，…，等其中的一個(gè)。其中，"*"是指任何一個(gè),"?"代表不確定,"0"代表沒有。
系統(tǒng)規(guī)定:夾具的類型，工件的形狀，位置模式不能是"*"和"?"。在設(shè)計(jì)系統(tǒng)時(shí),三個(gè)項(xiàng)目的屬性信息沒有這些選擇,這就意味著必須選擇確定的屬性。
最后三位數(shù)是事例識(shí)別號(hào)碼,如果事例特征的13位數(shù)是一樣的,這三個(gè)數(shù)字就用來(lái)區(qū)別他們。
該系統(tǒng)還規(guī)定:"000"是用于修正的一個(gè)典型事例,其他事例"001"、"002"、…,這些是用于設(shè)計(jì)師查找參考事例的. 如果其中一個(gè)偶爾需要改變成典型事例,首先它必須要求改成"000",前面的自動(dòng)變成參考事例.
事例索引碼的結(jié)構(gòu)如圖1所示。
1—夾具類型； 6—工件重量； 11—夾緊模型；
2—工件形狀； 7—工件剛度； 12—夾具體；
3—工件材料； 8—加工內(nèi)容； 13—其他；
4—批 量； 9—過程所有物； 14到16—事例識(shí)別碼；
5—工件比例； 10—定位模型；
圖1 事例索引碼的結(jié)構(gòu)
1.2事例庫(kù)
事例庫(kù)由許多預(yù)定義的事例組成。事例的描述是基于事例推理的最重要的問題之一。所以由索引碼復(fù)合。
1.3 事例的層次化
夾具的結(jié)構(gòu)相似被認(rèn)為是整個(gè)夾具，成分和內(nèi)容相似。所以，整個(gè)夾具事例庫(kù)，成分事例庫(kù)，夾具的成分事例庫(kù)形成相同。整個(gè)夾具的設(shè)計(jì)資料通常是由工件資料和工件加工資料組成,這就意味著夾具的設(shè)計(jì)應(yīng)滿足特別功能的需求.全部夾具事例是由功能成分組成，它是用功能成分的名字和數(shù)字來(lái)進(jìn)行描述的。成分事例代表成員（成分功能和其他結(jié)構(gòu)成分，主要驅(qū)動(dòng)參數(shù)，數(shù)字，和它們的約束關(guān)系）。成分事例（夾具的最低層）是功能成分和和其他成分的結(jié)構(gòu)。在現(xiàn)代夾具設(shè)計(jì)中有很多參數(shù)化準(zhǔn)件和普通非標(biāo)準(zhǔn)件。所以成分事例圖書館應(yīng)記錄特殊參數(shù)和保持它們的方法。
2事例修改的策略
在基于事例的夾具設(shè)計(jì)中，最重要的是相似點(diǎn)的修改，這樣能有助于獲得最相似的事例，以及縮短適應(yīng)時(shí)間。根據(jù)夾具設(shè)計(jì)的需求，事例修改的策略使最接近的事例方法和知識(shí)指導(dǎo)結(jié)合起來(lái)。首先在深度上查找，然后在寬度上；知識(shí)指導(dǎo)策略意味著在來(lái)自客觀事物根源的知識(shí)規(guī)則上查找，這就要首先查找固定類型，然后查找工件的形狀，第三查找定位方法。例如，如果事例索引碼包括夾具類型的磨削夾具，就只查找所有的磨削夾具，然后查找工件形狀的盒子，第三查找一個(gè)平面兩個(gè)銷的定位方法。如果沒有合適的，就查找深度標(biāo)點(diǎn)，然后回到最上層，然后再找所有與寬度相關(guān)的事例。
修改方法：
1） 根據(jù)夾具事例庫(kù)的事例索引信息，查找有關(guān)事例庫(kù)。
2） 將事例索引碼與事例庫(kù)的每個(gè)事例碼匹配，然后計(jì)算相似尺寸的價(jià)值。
3） 整理相似尺寸的次序，最大的架子是最類似的事例。
兩個(gè)事例之間的相似點(diǎn)是基于兩個(gè)事例特征之間的相似點(diǎn)。相似點(diǎn)尺寸的計(jì)算依靠特征的類型。相似點(diǎn)的價(jià)值可以通過數(shù)字化的價(jià)值來(lái)計(jì)算，例如比較重量分別是50kg 和 20kg的工件。非數(shù)字化的價(jià)值也能計(jì)算，例如，現(xiàn)在前13位索引碼都是非數(shù)字化的價(jià)值。一個(gè)夾具的相似尺寸的計(jì)算公式如下：
其中S表示通用夾具的相似尺寸，n表示索引特性數(shù)，表示每個(gè)特性的重量，表示事例庫(kù)中特性和相關(guān)夾具的特性的相似尺寸。同時(shí)， ，數(shù)值計(jì)算如下：
其中表示第i個(gè)特征的索引特性值，表示事例庫(kù)中第j個(gè)事例的第i個(gè)特征的特性值。
所以有兩種方法選擇相似夾具。一個(gè)方法是建立數(shù)值。如果通用事例的相似尺寸值比給定的數(shù)值小，這些事例就不能選來(lái)作相似事例。事例庫(kù)最初建立的時(shí)候，只有一些事例，數(shù)值可以建小一點(diǎn)。如果有大量的相似事例，數(shù)值就應(yīng)該建的大一些。另外一個(gè)方法是只建立相似事例的數(shù)字（例如10），這是類型單里相似尺寸的最大值。
3 事例的修改和存儲(chǔ)
3.1事例的修改
夾具設(shè)計(jì)中相似事例的修改包括以下三個(gè)階段：
1） 成分的替代
2） 保持形式不變，調(diào)整成分的特性
3） 模型重新設(shè)計(jì)
如果夾具的成分是普通的物品，它們能通過使用工具被修改，代替以及刪除，這些已經(jīng)被設(shè)計(jì)好了。
3.2事例的存儲(chǔ)
在將一個(gè)新的事例保存到事例庫(kù)之前，設(shè)計(jì)者必須考慮保存是否有價(jià)值。如果這個(gè)事例不能增加系統(tǒng)的知識(shí)，就沒有必要把它保存到事例庫(kù)里。如果它有價(jià)值的話，設(shè)計(jì)者在保存之前必須分析一下，看看這個(gè)事例是否作為標(biāo)準(zhǔn)事例或參考事例被存儲(chǔ)了。一個(gè)標(biāo)準(zhǔn)事例是一個(gè)描述同族事例主要特征的標(biāo)準(zhǔn)。一個(gè)同族事例是有事例庫(kù)中索引碼前13位相同而最后三位不同的那些事例組成的。一個(gè)標(biāo)準(zhǔn)事例的最后三位通常是“000”。一個(gè)參考事例屬于同族標(biāo)準(zhǔn)事例，最后三位用不同數(shù)字區(qū)分。
從被解釋的概念中，可采用以下方法：
1）如果一個(gè)新的事例和任何一個(gè)存在的事例族一致，和一個(gè)存在的標(biāo)準(zhǔn)事例的前13位數(shù)相同，那么這個(gè)事例就不能存儲(chǔ)因?yàn)橐呀?jīng)這種標(biāo)準(zhǔn)事例了?；蛘咧荒茏鳛橐粋€(gè)參考事例保存（最后三位不是“000”，而且和其它的不一樣）在事例庫(kù)中。
2）如果一個(gè)新的事例和任何一個(gè)存在的事例族一致，并且被認(rèn)為代替這個(gè)事例族要比以前的標(biāo)準(zhǔn)事例好，那么這個(gè)標(biāo)準(zhǔn)事例就被這個(gè)新的事例代替，以前的標(biāo)準(zhǔn)事例作為一個(gè)參考事例保存。
3）如果一個(gè)新的事例和任何一個(gè)存在的事例族不一致，一個(gè)新的事例族將會(huì)自動(dòng)產(chǎn)生，并作為標(biāo)準(zhǔn)事例保存到事例庫(kù)中。
4夾具設(shè)計(jì)中基于事例推理的過程
根據(jù)夾具設(shè)計(jì)的特性，夾具設(shè)計(jì)的基本信息，例如夾具的名字，零件，生產(chǎn)和設(shè)計(jì)者等等，必須先輸入。然后，輸入或設(shè)計(jì)工件的模型。輸入有關(guān)工件的細(xì)節(jié)信息，建立事例索引碼，然后CBR開始依靠相似尺寸查找相似事例，選出最相似的事例。如果需要的話，事例要滿足通用性設(shè)計(jì)，再存儲(chǔ)到事例庫(kù)中。程序流程圖如圖2所示
圖2 基于事例推理的夾具設(shè)計(jì)流程圖
5基于事例推理的夾具設(shè)計(jì)說(shuō)明
這是一個(gè)工件如圖3所示。材料是45鋼，底座，形狀為塊狀，生產(chǎn)批量為中批等。需要設(shè)計(jì)成一個(gè)用來(lái)旋轉(zhuǎn)孔的旋轉(zhuǎn)夾具。
圖3 需要設(shè)計(jì)夾具的一個(gè)工件
（最大尺寸80mmx49mmx22mm）
工件的特征值，屬性值，事例索引碼和重量在表1中列出。
表1 工件的事例索引碼和重量
特征名稱 特性值 索引碼 重量
夾具類型 車床夾具 1 100
工件形狀 塊狀 9 90
工件材料 中碳鋼 3 70
批量 中批 2 60
工件比例 小 5 60
工件重量 輕 5 60
工件剛度 硬度強(qiáng) 1 60
加工內(nèi)容 孔 3 80
程序要求 完成加工 3 70
定位方法 三個(gè)平面 1 100
夾緊方法 不確定 ？ 90
夾具體 復(fù)合 4 80
其他 沒有 0 60
通過查找和計(jì)算相似點(diǎn)，最相似的事例的事例索引碼是19325513321402000，細(xì)節(jié)信息在表2中列出。
表2 最相似事例的事例索引碼
特征名稱 特性值 索引碼
夾具類型 車床夾具 1
工件形狀 塊狀 9
工件材料 中碳鋼 3
批量 中批 2
工件比例 小 5
工件重量 輕 5
工件剛度 硬度強(qiáng) 1
加工內(nèi)容 孔 3
程序要求 完成加工 3
定位方法 三個(gè)平面 1
夾緊方法 不確定 ？
夾具體 復(fù)合 4
其他 沒有 0
相似點(diǎn)的計(jì)算如下：
所以?shī)A具的相似尺寸值是0.806，這是在事例庫(kù)中用于設(shè)計(jì)的最相似的事例，最相似的事例的結(jié)構(gòu)如圖4所示
圖4 最相似的夾具
當(dāng)成分替代，修改定位模型和夾緊模型，以及調(diào)節(jié)相關(guān)尺寸之后，新的夾具被設(shè)計(jì)出來(lái)，圖形如圖6所示
圖5 需要設(shè)計(jì)的新夾具
因?yàn)樵谑吕龓?kù)中沒有相似夾具，新夾具被儲(chǔ)存到事例庫(kù)中。事例索引碼是19325523311402000。
6 結(jié)論
基于事例推理，作為一個(gè)問題解決的方法，是一個(gè)比模仿人類思想的專業(yè)系統(tǒng)更有效的方法，已經(jīng)在很多難獲取知識(shí)的領(lǐng)域里得到發(fā)展?；谑吕评淼膬?yōu)點(diǎn)如下：它和人類的思想很相似；一個(gè)事例庫(kù)通過保存新事例獲得自學(xué)能力，它比有慣例庫(kù)更快更容易，它可以更好的傳遞和解釋新的知識(shí)，這和慣例庫(kù)有很大的不同。基于事例推理中提出的一個(gè)夾具設(shè)計(jì)的框架已經(jīng)被實(shí)行了，使用的是支持基礎(chǔ)數(shù)據(jù)的VC++，UG電腦繪圖軟件。這個(gè)框架也已經(jīng)和普通成分庫(kù)和典型夾具庫(kù)結(jié)合起來(lái)。這個(gè)發(fā)展的標(biāo)準(zhǔn)系統(tǒng)，用于航空項(xiàng)目，幫助夾具設(shè)計(jì)者提高設(shè)計(jì)效率和重新使用先前的設(shè)計(jì)資源。
9
1ApplicationApplicationApplicationApplication andandandand developmentdevelopmentdevelopmentdevelopmentOfOfOfOf casecasecasecase basedbasedbasedbasedreasoningreasoningreasoningreasoning ininininfixturefixturefixturefixturedesigndesigndesigndesignAbstract:Abstract:Abstract:Abstract: Based on the case baseddesigning(CBD)methodology,thefixture similarity is in two respects: thefunction and the structure information.Then, the computer aided fixture designsystemiscreatedoncasebasedreasoning (CBR),in which the attributesof the main features of workpiece andstructure of fixture as case index codeare designed for the retrieve of thesimilar cases, and the structure andhierarchical relation of case library aresetupforstore.Meanwhile,thealgorithmbasedontheknowledgeguided in the retrieve of the similarcases, the strategy of case adapt at ionand case storage in which the case identif cat ion number is used to distinguishfrom similar cases are presented. Theapplicationofthesysteminsomeprojects improves the design efficiencyand gets a good result .Keywords:Keywords:Keywords:Keywords:casebasedreasoning;fixturedesign;computeraided design(CAD)Fixtures are devices that serveas the purpose of holding the workpiecesecurelyandaccurately,andmaintaining a consistent relationshipwithrespecttothetoolswhilemachining.Becausethefixturestructure depends on the feature of theproduct and the status of the processplanning in the enterprise, its design isthe bottleneck during manufacturing,whichrestrainstoimprovetheefficiency and leadtime. And fixture2design is a complicated process, basedonexperiencethatneedscomprehensive qualitative knowledgeaboutanumberofdesignissuesincludingworkpiececonfiguration,manufacturing processes involved, andmachining environment. This is also avery time consuming work when usingtraditionalCADtools(suchasUnigraphics, CATIA or Pro/E), whichare good at performing detailed designtasks, but provide few benefits fortaking advantage of the previous designexperience and resources, which areprecisely the key factors in improvingthe efficiency. The methodology of casebasedreasoning(CBR)adaptsthesolution of a previously solved case tobuild a solution for a new problem withthe following four steps: retrieve, reuse,revise, and retain 1. This is a moreuseful method than the use of an expertsystemtosimulatehumanthoughtbecause proposing a similar case andapplying a few modifications seems tobe self explanatory and more intuitivetohumans.Sovariouscasebaseddesignsupporttoolshavebeendevelopedfornumerousareas2-4,such as in injection molding and design,architectural design, die castingdiedesign, process planning, and also infixture design. Sun used six digitals tocompose the index code that includedworkpieceshape,machineportion,bushing, the 1stlocating device, the 2ndlocating device and clamping device5.But the system cannot be used for otherfixture types except for drill fixtures,and cannot solve the problem of storageof the same index code that needs to beretained, which is very important inCBR6.1 Construction of a Case Index and3Case Library1.1 Case indexThecaseindexshouldbecomposedofallfeaturesoftheworkpiece,whicharedistinguishedfrom different fixtures. Using all ofthem wouldmake the operation inconvenient. Because the forms of theparts are diverse, and the technologyrequirementsofmanufacture intheenterprise also develop continuously,lots of features used as the case indexwill make the search rate slow, and themainfeatureunimportant,forthereason that the relative weight which isallotted to every feature must diminish.And on the other hand, it is hard toinclude all the features in the caseindex.Therefore,consideringthepracticality and the demand of rapiddesign, the case index includes both themajor feature of the workpiece and thestructure of fixture. The case index codeis made up of 16 digits: 13 digits forcase features and 3 digits for caseidentification number.The first 13 digits represent 13features. Each digit is corresponding toan attribute of the feature, which maybe one of“*”, “？”, “1”, “2”,“A” , “B” , “Z” , etc. In which,“*” means anyone, “?” uncertain,“0” nothing.The system rules: fixture type,workpiece shape, locating model cannotbe “*”or“?”. When the system isdesigned, the attribute information ofthe three items does not have theseoptions,whichmeansthecertainattribute must be selected.The last three digits are the caseidentification number, which means the13 digits of the case feature are the4same, and the number of these threedigits is used for distinguishing them.The system also rules: “000”is aprototypecase,whichisusedforretrieval, and other cases are “001”,“ 002 ” ,which are used forreferencecasestobesearchedbydesigners. If occasionally one of themneeds to be changed as the prototypecase, first it must be required to applyto change the one to “000”, and theformer is changed to referential caseautomatically.The construction of the case indexcode is shown in Fig.1.1.2 Case libraryThe case library consists of lots ofpredefined cases. Case representation isone of the most important issues in casebased reasoning. So compounding withthe index code,.1.3 Hierarchical form of CaseThestructuresimilarityofthefixture is representedas the wholefixture similarity, components similarityand component similarity. So the wholefixture case library, components caselibrary,componentcaselibraryoffixtureareformedcorrespondingly.Usuallydesigninformationofthewhole fixture is composed of workpieceinformation and workpiece procedureinformation, which represent the fixturesatisfyingthespecificallydesigningfunction demand. The whole fixturecase is made up of function components,which are described by the functioncomponentsnames and numbers. Thecomponentscaserepresentsthemembers.(functioncomponentand5other structure components ，maindriven parameter, the number, and theirconstrain relations.)The componentcase (the lowest layer of the fixture) isthe structure of function component andothercomponents.Inthemodernfixturedesigntherearelotsofparametric standard parts and commonnon standard parts. So the componentcaselibraryshouldrecordthespecification parameter and the way inwhich it keeps them.2 Strategy of Case RetrievalInthecasebaseddesignoffixtures ,the most important thing is theretrieval of the similarity, which canhelp to obtain the most similar case, andto cut down the time of adaptation.According to the requirement of fixturedesign, the strategy of case retrievalcombinesthewayofthenearestneighbor and knowledge guided. That is,first search on depth, then on breadth;the knowledge guided strategy means tosearch on the knowledge rule from rootto the object, which is firstly searchedby the fixture type, then by the shape ofthe workpiece, thirdly by the locatingmethod. For example, if the case indexcode includes the milling fixture offixture type, the search is just for allmillingfixtures,thenforboxofworkpiece shape, the third for 1plane+2pine of locating method. If there is nomatch of it, then the search stops ondepth, and returns to the upper layer,and retrieves all the relative cases onbreadth.Retrieval algorithms:1)Accordingtothecaseindexinformationoffixturecaselibrary,search the relevant case library;2)Match the case index code withthe code of each case of the case library,6and calculate the value of the similaritymeasure;3)Sorttheorderofsimilaritymeasure, the biggest value, which is themost analogical case.Similarity between two cases isbased on the similarity between the twocases.features.Thecalculationofsimilarity measure depends on the typeof the feature. The value of similaritycan be calculated for numerical values,for example, compareWorkpiece withthe weight of 50kg and 20kg. The valuecan also be calculated between nonnumerical values, for example, now thefirst 13 digits index code is all nonnumericalvalues.Thesimilaritymeasure of a fixture is calculated asfollows:where S is the similarity measureof current fixture, n is the number of theindex feature,is the weight of eachfeature,is the similaritymeasure of the attributeof the i2thfeature with the attributeof relativefeature of the j-th case in the caselibrary.Atthesametime, the value countsas follows:.Whereis the value of the indexattribute of the i-th feature, andis7the value of attribute of the relative i-thfeature of the j-th case in case library.So there are two methods to selectthe analogical fixture. One is to set thevalue.Ifthevaluesofsimilaritymeasure of current cases were less thana given value, those cases would not beselected as analogical cases. When thecase library is initially set up, and thereare only a few cases, the value can besetsmaller.Iftherearelotsofanalogical cases, the value should getlarger. The other is just to set thenumber of the analogical cases (suchas10), which is the largest value ofsimilaritymeasurefromthesortedorder.3CaseadaptationandCaseStorage3.1 Case adaptationThemodificationoftheanalogical case in the fixture designincludes the following three cases:1) The substitution of componentsand the component;2) Adjustingthedimensionofcomponents and the component whilethe form remains;3) The redesign of the model.If the components and componentof the fixture are common objects, theycan be edited, substituted and deletedwith tools, which have been designed.3.2 Case storageBefore saving a new fixture case inthecase library, the designermustconsider whether the saving is valuable.Ifthecasedoesnotincreasetheknowledge of the system, it is notnecessary to store it in the case library.If it is valuable, then the designer mustanalyzeitbefore savingittoseewhetherthecaseisstoredasaprototype case or as reference case. A8prototype case is a representation thatcan describe the main features of a casefamily. A case family consists of thosecases whose index codes have the samefirst 13 digits and different last threedigits in the case library. The last threedigits of a prototype case are always“000” . A reference case belongs to thesame family as the prototype case andis distinguished by the different lastthree digits.From the concept that has beenexplained, the following strategies areadopted:1) If a new case matches anyexisting case family, it has the samefirst 13 digits as an existing prototypecase, so the case is not saved because itis represented well by the prototypecase. Or is just saved as a referencecase (the last 3 digits are not “000”,and not the same with others) in thecase library.2) If a new case matches anyexisting case family and is thought tobebetteratrepresentingthiscasefamily than the previous prototype case,then the prototype case is substituted bythisnewcase,andthepreviousprototype case is saved as a referencecase.3) If a new case does not matchany existing case family, a new casefamily will be generated automaticallyand the case is stored as the prototypecase in the case library.4ProcessofCBRinFixtureDesignAccording to the characteristics offixture design, the basic information ofthe fixture design such as the name offixture, part, product and the designer,etc. must be input first. Then the fixturefile is set up automatically, in which all9componentsofthefixtureareputtogether.Thenthemodeloftheworkpiece is input or designed. Thedetailedinformationabouttheworkpiece is input, the case index codeis set up, and then the CBR begins tosearch the analogical cases, relying onthe similarity measure, and the mostanalogicalcaseisselectedout.Ifneeded, the case is adapted to satisfythe current design, and restored into thecaselibrary.Theflowchartoftheprocess is shown in Fig.3.5 Illustrating for Fixture Design byCBRThis is a workpiece (seeFig.4). Itsmaterial is 45# steel. Its name is seat.Its shape is block, and the product batchsize is middle, etc.Afixture is turningfixture that serves to turn the hole,which needs to be designed.The value of feature, attribute,case index code and weight of theworkpiece is show n in Tab.2.10Through searching, and calculatingthe similarity, the case index code of themostsimilarcaseis19325513321402000, and the detailedinformation is show n in Tab. 3.The similarity is calculated asfollows:So the value of similarity measureofthefixturewhichneedstobedesigned with the most analogical caseincaselibraryis0.806,andthestructure of the most analogical case isshown in Fig.5.After having been substituted thecomponent,modifiedthelocatingmodel and clamp model, and adjustedthe relative dimension, the new fixtureis designed, and the figure is show n in11Fig.6.As there is not the analogicalfixture in the case library, the newfixture is restored in to the case library.Thecaseindexcodeis19325513311402000.6 ConclusionCBR,asaproblemsolvingmethodology,isamoreefficientmethodthananexpertsystemtosimulate human thought, and has beendevelopedinmanydomainswhereknowledge is difficult to acquire. Theadvantages of the CBR are as follows:itresembleshumanthoughtmoreclosely; the building of a case librarywhichhasselflearningabilitybysaving new cases is easier and fasterthan the building of a rule library; and itsupportsabettertransferandexplanation of new knowledge that ismore different than the rule library.Aproposed fixture design framework onthe CBR has been implemented byusing Visual C +, UG/Open API in Un graphics with Oracle as databasesupport, which also has been integratedwiththe32Dparametriccommoncomponentlibrary,commoncomponents library and typical fixturelibrary.Theprototypesystem,developed here, is used for the aviationproject, and aids the fixture designers toimprove the design efficiency and reuseprevious design resources.
收藏