本文選題：設(shè)計(jì)知識(shí)管理 + 知識(shí)重用�。� 參考：《北京理工大學(xué)》2014年博士論文

【摘要】：產(chǎn)品設(shè)計(jì)是產(chǎn)品開發(fā)過程中的創(chuàng)造性階段，其本質(zhì)是一個(gè)基于既有知識(shí)產(chǎn)生新知識(shí)的過程。在產(chǎn)品設(shè)計(jì)領(lǐng)域大約70%的產(chǎn)品設(shè)計(jì)為變型設(shè)計(jì)，并且在新產(chǎn)品設(shè)計(jì)時(shí)大約60%以上的工作需要以往經(jīng)驗(yàn)的支撐。因此如何管理與重用已有的設(shè)計(jì)知識(shí)成為提升產(chǎn)品設(shè)計(jì)效率和質(zhì)量的關(guān)鍵問題。然而，目前設(shè)計(jì)知識(shí)管理與重用領(lǐng)域的研究工作主要集中在設(shè)計(jì)知識(shí)的表示、建模以及通用的知識(shí)服務(wù)技術(shù)方面，，對(duì)設(shè)計(jì)人員的知識(shí)重用模式以及相應(yīng)的個(gè)性化知識(shí)服務(wù)方法的研究仍然不足。這種不足直接導(dǎo)致了設(shè)計(jì)知識(shí)重用時(shí)產(chǎn)生嚴(yán)重的信息過載問題，也致使設(shè)計(jì)知識(shí)處于“管而不用”的狀態(tài)。針對(duì)以上問題，論文以提升設(shè)計(jì)知識(shí)的重用水平、減輕信息過載為目標(biāo)，從理論和方法層面系統(tǒng)性地研究了設(shè)計(jì)知識(shí)導(dǎo)航、設(shè)計(jì)知識(shí)檢索以及設(shè)計(jì)知識(shí)推送等關(guān)鍵技術(shù)，為滿足設(shè)計(jì)過程中的個(gè)性化知識(shí)服務(wù)需求提供了技術(shù)支撐，采用試驗(yàn)和案例的方法驗(yàn)證了技術(shù)的有效性。論文的主要研究內(nèi)容如下： （1）設(shè)計(jì)知識(shí)管理與重用技術(shù)框架研究。針對(duì)設(shè)計(jì)知識(shí)管理系統(tǒng)難以滿足設(shè)計(jì)人員知識(shí)需求的問題，首先研究并歸納出了設(shè)計(jì)人員的四種知識(shí)重用模式，包括參考、導(dǎo)引、自動(dòng)以及協(xié)助；在分析知識(shí)重用模式的基礎(chǔ)上，研究了設(shè)計(jì)知識(shí)管理與重用系統(tǒng)在整個(gè)設(shè)計(jì)支撐體系中的定位并分析了該系統(tǒng)與其他系統(tǒng)的數(shù)據(jù)和功能交互關(guān)系；最后從設(shè)計(jì)知識(shí)、設(shè)計(jì)人員以及系統(tǒng)架構(gòu)三個(gè)角度研究了設(shè)計(jì)知識(shí)管理與重用系統(tǒng)的技術(shù)需求。該研究內(nèi)容為論文的研究提供了總體框架。 （2）異構(gòu)設(shè)計(jì)知識(shí)建模與組織方法研究。針對(duì)設(shè)計(jì)知識(shí)種類繁多、來源廣泛、存儲(chǔ)方式異構(gòu)導(dǎo)致的難以實(shí)現(xiàn)統(tǒng)一管理機(jī)制和統(tǒng)一語義框架的問題，提出了面向異構(gòu)設(shè)計(jì)知識(shí)的多層次組織方法。該方法分三個(gè)層次組織設(shè)計(jì)知識(shí)，在底層研究了基于知識(shí)組件的設(shè)計(jì)知識(shí)建模方法來實(shí)現(xiàn)設(shè)計(jì)知識(shí)的封裝和快速重用；在中層構(gòu)建了設(shè)計(jì)知識(shí)統(tǒng)一模型，該模型通過提取設(shè)計(jì)知識(shí)的元信息實(shí)現(xiàn)設(shè)計(jì)知識(shí)的結(jié)構(gòu)一致化；在上層研究了設(shè)計(jì)知識(shí)與領(lǐng)域本體的映射方法，實(shí)現(xiàn)了設(shè)計(jì)知識(shí)語義一致化。 （3）領(lǐng)域設(shè)計(jì)知識(shí)導(dǎo)航方法研究。針對(duì)跨領(lǐng)域設(shè)計(jì)知識(shí)重用時(shí)設(shè)計(jì)人員獲取知識(shí)的盲目性，提出了基于知識(shí)地圖的領(lǐng)域知識(shí)導(dǎo)航方法。首先研究了基于自然語言處理技術(shù)、潛在語義分析技術(shù)以及相似度分析技術(shù)自動(dòng)化構(gòu)建領(lǐng)域知識(shí)地圖的方法；其次研究并提出了基于復(fù)雜網(wǎng)絡(luò)理論的知識(shí)地圖內(nèi)部結(jié)構(gòu)探析以及提取關(guān)鍵領(lǐng)域知識(shí)的方法；最后研究了利用知識(shí)地圖和關(guān)鍵領(lǐng)域知識(shí)對(duì)設(shè)計(jì)人員進(jìn)行知識(shí)導(dǎo)航的方法。實(shí)驗(yàn)證明該方法能夠提供有效的知識(shí)導(dǎo)航路徑，使設(shè)計(jì)人員快速地重用跨領(lǐng)域的設(shè)計(jì)知識(shí)。 （4）個(gè)性化設(shè)計(jì)知識(shí)檢索方法研究。針對(duì)傳統(tǒng)基于內(nèi)容的知識(shí)檢索結(jié)果不能體現(xiàn)用戶偏好，容易造成信息過載的問題，提出了基于多行為的個(gè)性化知識(shí)檢索方法。首先研究并構(gòu)建了設(shè)計(jì)知識(shí)間行為相似度的計(jì)算模型，該模型綜合考慮了行為表征用戶偏好的能力、行為之間的依賴關(guān)系以及用戶的部門隸屬關(guān)系三個(gè)方面的因素；在此基礎(chǔ)上研究了基于知識(shí)行為相似度的知識(shí)檢索結(jié)果過濾方法。實(shí)驗(yàn)證明基于該模型實(shí)現(xiàn)的知識(shí)檢索方法明顯改善了傳統(tǒng)知識(shí)檢索的效能，有助于解決設(shè)計(jì)人員的信息過載問題。 （5）設(shè)計(jì)知識(shí)推送方法研究。針對(duì)產(chǎn)品設(shè)計(jì)知識(shí)復(fù)雜多樣、語義信息難以獲取所造成的重用難題，提出了基于概率的混合用戶模型以及知識(shí)推送方法。首先研究了將用戶評(píng)分與知識(shí)主題相結(jié)合的可行性判定準(zhǔn)則；其次將用戶興趣定義為對(duì)特定主題給出某一評(píng)分的概率，并綜合了基于行為的方法和基于內(nèi)容的方法的優(yōu)勢(shì)構(gòu)建混合用戶模型；最后將該模型與經(jīng)典的協(xié)同過濾算法結(jié)合實(shí)現(xiàn)了設(shè)計(jì)知識(shí)推送方法。試驗(yàn)證明該方法在預(yù)測(cè)準(zhǔn)確性和覆蓋率方面都優(yōu)于其他應(yīng)用相同信息進(jìn)行知識(shí)推送的方法。 （6）作為上述研究成果的工程實(shí)踐，自主設(shè)計(jì)并開發(fā)了產(chǎn)品設(shè)計(jì)知識(shí)管理與重用系統(tǒng)。該系統(tǒng)以個(gè)性化的方式提供了知識(shí)檢索、知識(shí)推送以及知識(shí)導(dǎo)航等知識(shí)服務(wù)功能，支持設(shè)計(jì)人員在設(shè)計(jì)系統(tǒng)中直接獲取所需要的設(shè)計(jì)知識(shí)。該系統(tǒng)所提供的知識(shí)服務(wù)能夠轉(zhuǎn)變?cè)O(shè)計(jì)知識(shí)“管而不用”的現(xiàn)狀，同時(shí)顯著降低信息過載現(xiàn)象的發(fā)生。
[Abstract]:Product design is the creative stage in the process of product development , its essence is a process of generating new knowledge based on the existing knowledge . The research work in the field of design knowledge management and reuse is a key problem to improve the efficiency and quality of product design .

( 1 ) Design knowledge management and reuse technical framework research . In view of the difficulty of designing knowledge management system to meet the knowledge requirement of designers , the four kinds of knowledge reuse patterns , including reference , guidance , automatic and assistance , are researched and summarized .
Based on the analysis of the pattern of knowledge reuse , the paper studies the location of the design knowledge management and reuse system in the whole design support system and analyzes the data and function interaction relationship between the system and other systems ;
Finally , the technical requirements of the design knowledge management and reuse system are studied from three aspects : design knowledge , designer and system architecture .

( 2 ) Research on the modeling and organization method of heterogeneous design knowledge . Aiming at the problem of difficult realization of unified management mechanism and unified semantic framework , which is caused by many kinds of design knowledge , wide source and heterogeneous storage mode , a multi - level organization method for heterogeneous design knowledge is proposed .
In the middle layer , a unified model of design knowledge is constructed , which is consistent with the structure of design knowledge by extracting meta - information of design knowledge ;
The mapping method of design knowledge and domain ontology is studied in the upper layer , and the semantic consistency of design knowledge is realized .

( 3 ) Research on the knowledge navigation method in the field of design knowledge . Based on the blindness of knowledge in cross - domain design knowledge reuse , a domain knowledge navigation method based on knowledge map is proposed . Firstly , a knowledge map method based on natural language processing technology , potential semantic analysis technology and similarity analysis technology is studied .
Secondly , the internal structure of knowledge map based on complex network theory and the methods of extracting knowledge in key areas are put forward .
Finally , the method of knowledge navigation using knowledge map and key field knowledge is studied . The experiment proves that the method can provide effective knowledge navigation path , so that the designer can quickly reuse the design knowledge in cross - domain .

( 4 ) The research on the retrieval method of personalized design knowledge . Aiming at the problem that the traditional content - based knowledge retrieval results do not reflect the user ' s preference and easily cause information overload , a multi - action - based personalized knowledge retrieval method is proposed .
On the basis of this , a knowledge retrieval result filtering method based on knowledge behavior similarity is studied . The experiment proves that the knowledge retrieval method based on the model improves the efficiency of traditional knowledge retrieval , and helps to solve the problem of information overload of designers .

( 5 ) Design knowledge push method research . Aiming at the complex variety of product design knowledge and difficult acquisition of semantic information , a hybrid user model based on probability and knowledge pushing method are put forward .
Secondly , the user interest is defined as the probability of giving a certain score to a particular topic , and the mixed user model is constructed based on the advantages of the behavior - based method and the content - based method ;
Finally , the design knowledge pushing method is implemented by combining the model with the classical collaborative filtering algorithm . The test proves that the method is superior to other methods for knowledge push in terms of accuracy and coverage .

( 6 ) As the engineering practice of the above research results , the product design knowledge management and reuse system is designed and developed . The system provides the knowledge service functions such as knowledge retrieval , knowledge push and knowledge navigation in a personalized way , and supports the design personnel to directly acquire the required design knowledge in the design system . The knowledge service provided by the system can change the present situation of the design knowledge " pipe instead of " while significantly reducing the occurrence of the information overload phenomenon .
【學(xué)位授予單位】：北京理工大學(xué)
【學(xué)位級(jí)別】：博士
【學(xué)位授予年份】：2014
【分類號(hào)】：TB472

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 鄧志鴻,唐世渭,張銘,楊冬青,陳捷;Ontology研究綜述[J];北京大學(xué)學(xué)報(bào)(自然科學(xué)版);2002年05期

2 陳永當(dāng);楊海成;杜兵勞;;基于本體的航空發(fā)動(dòng)機(jī)設(shè)計(jì)知識(shí)組織模型構(gòu)建與分析[J];航空動(dòng)力學(xué)報(bào);2007年01期

3 林慧蘋;范玉順;黃琛;;產(chǎn)品設(shè)計(jì)中基于元模型的知識(shí)管理[J];計(jì)算機(jī)集成制造系統(tǒng);2007年04期

4 郝佳;楊海成;閻艷;王國新;;面向產(chǎn)品設(shè)計(jì)任務(wù)的可配置知識(shí)組件技術(shù)[J];計(jì)算機(jī)集成制造系統(tǒng);2012年04期

5 于海濤;高一波;楊一平;;基于知識(shí)樹的領(lǐng)域知識(shí)組織和應(yīng)用[J];計(jì)算機(jī)應(yīng)用研究;2008年11期

6 藺豐奇;劉益;;信息過載問題研究述評(píng)[J];情報(bào)理論與實(shí)踐;2007年05期

7 賈磊;洪軍;邱志惠;李寶童;閻海紅;潘世翼;;復(fù)雜結(jié)構(gòu)零件的初始化設(shè)計(jì)知識(shí)模型構(gòu)建[J];西安交通大學(xué)學(xué)報(bào);2008年07期

8 ;CONSTRUCTION METHOD OF KNOWLEDGE MAP BASED ON DESIGN PROCESS[J];Chinese Journal of Mechanical Engineering;2007年03期

9 謝友柏;現(xiàn)代設(shè)計(jì)與知識(shí)獲取[J];中國機(jī)械工程;1996年06期

10 李濤;;工作過程中信息過載的調(diào)查與分析研究[J];中國軟科學(xué);2003年09期

本文編號(hào)：2064862