【作者】 杨洁；

【导师】 杨育；

【作者基本信息】 重庆大学 ， 工业工程， 2009， 博士

【摘要】 近年来,客户知识已经被证明是企业创新最重要的资源之一。在协同产品创新过程中,如何围绕企业创新的要求,对丰富的客户信息进行分析、整理,提取协同创新所需的客户知识,并将其与产品创新方法相融合,从而辅助和支持创新团队的决策,提升企业的创新能力,已成为企业在市场竞争中获得有利地位的重要途径。本文在系统研究国内外客户知识管理、协同产品设计以及创新方法最新理论成果的基础上,针对网络环境下协同产品创新中客户知识集成模式及其关键技术进行了深入研究。全文的研究内容主要包括以下几个部分:首先,论文研究了协同产品创新中客户知识集成模式。为了使协同产品创新中客户知识的集成更能符合企业创新的要求,针对客户知识在产品协同创新中的模糊性、动态性、多样性、分布性等特点,结合协同产品创新的过程,建立了包括数据源层、信息层、知识层与集成应用层在内的协同创新环境下的客户知识集成模型。对协同产品创新中的人因、组织、环境、技术以及过程等约束条件进行了深入分析,研究了基于协调理论和Petri Nets的客户知识集成过程,以及协同环境中客户知识集成范式。同时,对协同产品创新中客户知识资源进行识别和评价,从五个方面建立了包括学习效应在内的客户知识源综合评价指标体系。为了减少冗余指标项运用粗糙集方法对评价指标进行了预处理并降低了小波网络的输入维数。采用迭代梯度下降法和逐步检验法确定小波网络结构基础上,进行了协同产品创新中客户知识源的综合评价。其次,论文研究了网络协同环境中客户信息发现及知识获取的方法。为了从大量、不完全、有噪声、模糊的数据中提取出可信、有效、可理解的客户信息,并将其与产品特征联系以支持协同产品创新活动,研究了基于协作过滤的客户偏好信息获取方法。同时根据客户满意度及最大化技术特征表现度进行客户需求的质量屋配置,采用模糊度量质量屋进行特征映射,以实现客户需求信息的获取。提出一种基于认知理论的客户感性知识获取方法,以认知行为为标准,利用特征匹配进行客户感性知识的识别,建立了基于模糊认知图的客户感性认知与产品特征的映射模型,利用蚁群算法确定模糊认知图的结构及邻接矩阵,从而实现对客户感性知识的获取。提出了协同产品创新中基于本体理论的客户知识建模方法。分析了客户知识建模的需求,并定义了客户知识建模中的概念。为了实现在多种不同条件约束下,客户对产品创新中的问题进行描述、表达、转换或求解,提出了基于“问题域-功能域”的客户知识转换方法。同时使用本体理论实现客户知识的建模,构建了包括“客户创意、感性知识、客户需求、基本知识”在内的客户知识概念本体树,实现了对客户知识的本体描述。同时采用基于语义相似度的客户知识本体树匹配算法以及基于约束满足的知识推送方法来实现客户知识请求匹配。提出了基于TRIZ的客户协同产品创新方法。为了更好地将客户知识与协同产品创新方法有机融合,定义了客户知识生命周期(Customer Knowledge Lifecycle,CKL)和客户知识产品化(Customer Knowledge Production,CKP)的概念,在此基础上建立了面向客户知识利用的协同产品创新框架。然后分析了发明创新的TRIZ理论及基于TRIZ的协同产品创新过程,提出基于TRIZ的客户协同产品创新方法,包括:客户知识的描述转换、基于TRIZ的客户知识语义模型建立、基于三维空间的领域问题求解以及基于客户知识及最短路径的产品创新方案分析与评价。最后开发了基于本文研究成果的协同产品创新中客户知识集成支持原型系统。详细介绍了面向客户知识利用的协同产品创新系统框架及构成,在系统框架的基础上,介绍了基于Web Services技术的,实现该原型系统的知识集成技术方案,实现了对客户知识本体的具体描述。然后利用原型系统中的TRIZ工具和客户知识对摩托车车架进行创新,通过应用示例验证本文所提出理论与方法的正确性和实用性。更多还原

【Abstract】 Customer knowledge is a precious resource to produce innovative products. The necessity about how to improve competition capability and hold vantage point for enterprises claim to analyze customer information, find available knowledge, combine with production innovation method. By this means, the output of efforts will be beneficial to support innovation team’s decision and improve enterprise’s innovation capability. Therefore, research work focus on integration of customer knowledge (CK) and key technology in collaborative product innovation(CPI), based on comprehensive study on customer knowledge management, collaborative product design and innovation methodology.Firstly, the mode of customer knowledge integration in CPI is studied. Summative characters of customer knowledge, such as fuzziness, Dynamic, diversity and distributivity, and key constrains of knowledge integration including human factor, organization, environment, technology and process is studied to meet enterprise innovation requirements. A CK’s integration model including CK integration process based on Coordinate Theory (CT) and Petri Nets (PN), including data resource layer, information layer, knowledge layer and integration application layer, is proposed. Furthermore, customer resources identification and evaluation system is manifested mainly in five items taking the study effect as an example. The input dimension of Wavelet Networks, whose structure is ascertained by Gradient Iteration Descent Algorithm and Step-test Procedure Algorithm, is utilized to evaluate innovation customer, is decreased by Rough Set Algorithm by reducing redundant indexes.Secondly, customer information discovery and knowledge acquisition is studied in networked collaborative surroundings. Customer preference information acquisition based on collaboration filtration is proposed so that reliable, effective and accessible customer information is acheieved from mass, imperfect, noisy and fuzzy data. Meanwhile, it can be linked with product characters to support product design. A method for customer sensibility knowledge acquisition based on Cognitive Theory is proposed to identify it by character matching with cognitive behavior. And a mapping model between customer sensibility knowledge and product character using fuzzy cognitive chart assists to obtain customer cognitive knowledge. Ant Colony Algorithm is betaken to confirm the structure of fuzzy cognitive chart and adjacency matrix. Thirdly, method for customer knowledge modeling based on ontology theory is proposed in CPI with definition and requirements analysis. Customer knowledge based‘problem domain-function domain’method is presented serving as describing, expressing, transforming or solving problems in CPI in different constrains. Ontology Theory and OWL language are used to model and describe customer knowledge by customer knowledge conception ontology tree including‘customer originality, sensibility knowledge, customer requirement and fundamental knowledge’. Matching algorithm for customer knowledge conception ontology tree and knowledge push method based constrain satisfaction are proposed to meet customer knowledge requirement.Fourthly, the method of CPI for customer knowledge utilization is proposed. Customer Knowledge Lifecycle (CK) and Customer Knowledge Production (CKP) are defined to combine customer knowledge and CPI method with CPI frame for knowledge utilization. CPI method based on TRIZ and customer knowledge are studied consisting of CK’s description transformation, CK’s grammar model, adjacency domain problem solving based on 3-D space and innovation evaluation based on CPM.Finally, Case study is demonstrated to illustrate the feasibility and applicability. A CPI prototype system for customer knowledge utilization is developed. Its frame and structure are presented including some key technologies such as Web Service. The prototype system’s knowledge integration solution and CK’s ontology description come to execution and support motor bicycle’s frame innovation design with the help of TRIZ tool and customer knowledge.更多还原