本文選題：知識建模 + 知識重用　； 參考：《浙江大學》2011年碩士論文

【摘要】：在當前以顧客為導向的市場環(huán)境下,多元化和個性化的客戶需求使得產(chǎn)品復雜性增加、設(shè)計周期縮短、產(chǎn)品更新加快,這給企業(yè)產(chǎn)品設(shè)計提出了很大的挑戰(zhàn)。如何對眾多產(chǎn)品知識進行高效地管理與利用,以提高產(chǎn)品設(shè)計的效率與質(zhì)量,是企業(yè)保證其競爭力的關(guān)鍵。針對國內(nèi)機械類企業(yè)尤其是中小企業(yè)中存在的設(shè)計知識管理混亂、產(chǎn)品開發(fā)周期長、知識重用率低等問題,本文將進行以下研究： 1.在歸納現(xiàn)有機械類產(chǎn)品設(shè)計知識的基礎(chǔ)上,提出了基于本體的知識表達,通過特征類概念、過程類概念、原理類概念以及實例類概念組成的四元本體對機械類產(chǎn)品知識進行表示,并給出了知識表示模型的搭建流程,從而使得機械類產(chǎn)品知識具有一個規(guī)范性的描述。 2.針對機械類產(chǎn)品的結(jié)構(gòu)特點,本文提出了組合特征的實例搜索策略,并采用最近相鄰算法與知識引導法相結(jié)合的方式進行實例相似度計算,通過基于K-均值的權(quán)值算法對特征相似度權(quán)值進行智能學習,實現(xiàn)產(chǎn)品相似實例快速準確的檢索。 3.利用Petri網(wǎng)對產(chǎn)品設(shè)計過程進行建模,針對傳統(tǒng)Petri網(wǎng)存在的缺點提出了兩點改進,一是提出了Petri塊的概念,用以縮小Petri網(wǎng)的規(guī)模,使得Petri網(wǎng)模型具有多層建模功能；二是對變遷加以謂詞模塊,以提升變遷的邏輯處理能力。在上述兩處改進的基礎(chǔ)上,提出了相應(yīng)的機械類產(chǎn)品設(shè)計過程的Petri網(wǎng)建模方法。 4.文章最后,基于原有的PDM系統(tǒng),以Visual Studio作為開發(fā)環(huán)境,SQL Server作為數(shù)據(jù)庫系統(tǒng),并用文中提出的知識建模方式開發(fā)摩托車制動器設(shè)計系統(tǒng)。
[Abstract]:In the current customer-oriented market environment, diversification and individuation of customer demand increase the complexity of products, shorten the design cycle, and accelerate product renewal, which brings a great challenge to enterprise product design. How to efficiently manage and utilize the knowledge of many products in order to improve the efficiency and quality of product design is the key for enterprises to ensure their competitiveness. In order to solve the problems of confused design knowledge management, long product development cycle and low knowledge reuse rate in domestic mechanical enterprises, especially in small and medium-sized enterprises, this paper will carry out the following research: 1. On the basis of summarizing the existing knowledge of mechanical product design, this paper puts forward the knowledge representation based on ontology, which is based on the concept of feature class, the concept of process class, the concept of feature class and the concept of process class. The concept of principle class and the concept of example class are composed of quaternion ontology to express the knowledge of mechanical product, and the process of building knowledge representation model is given, which makes the knowledge of mechanical product have a normative description. 2. According to the structural characteristics of mechanical products, this paper proposes a case search strategy based on combinatorial features, and uses the nearest neighbor algorithm and knowledge guidance method to calculate the similarity of cases. Based on the K- mean weight algorithm, the feature similarity weights are intelligently learned, and the product similarity examples can be retrieved quickly and accurately. 3. Petri net is used to model the product design process. Two improvements are put forward in view of the disadvantages of traditional Petri net. Firstly, the concept of Petri block is put forward to reduce the scale of Petri net and make the Petri net model have multi-layer modeling function. Second, predicate module is used to improve the logic processing ability of transition. Based on the above two improvements, a corresponding Petri net modeling method for mechanical product design process is proposed. 4. Finally, based on the original PDM system, using Visual Studio as the development environment, SQL Server is used as the database system, and the motorcycle brake design system is developed with the knowledge modeling method proposed in this paper.
【學位授予單位】：浙江大學
【學位級別】：碩士
【學位授予年份】：2011
【分類號】：TH122

【引證文獻】

相關(guān)期刊論文 前1條

1 陳君;;基于實例的汽車零部件產(chǎn)品開發(fā)中關(guān)鍵技術(shù)研究[J];湖北理工學院學報;2013年01期

相關(guān)碩士學位論文 前2條

1 張冬明;產(chǎn)品設(shè)計知識管理機制研究[D];天津大學;2012年

2 李榮鵬;液壓挖掘機設(shè)計專家系統(tǒng)研究[D];浙江大學;2012年

，

本文編號：1914730