【作者】 张军；

【导师】 赵江洪； 孙宗禹；

【作者基本信息】 湖南大学 ， 设计学， 2013， 博士

【摘要】 现代设计的重心已由传统的基于经验的设计转变为基于知识的设计，以数控机床为代表的复杂产品的造型设计为工业设计研究和实践带来了新的挑战和机遇，面向复杂产品的工业设计创新话题，论文提出了基于知识的辅助工业设计系统的概念并通过一系列的研究得以初步实现。论文研究的目标是面向数控机床这个典型复杂产品，开发基于知识的支持数控机床产品造型设计的辅助设计系统，以解决数控机床造型设计问题、满足装备产品行业信息化设计能力的提升。研究的主要内容是复杂产品工业设计知识的获取和在系统层面的表达和实现技术，以及基于知识的辅助工业设计系统平台的构建。论文研究在多项国家科技攻关和支撑计划的支持下，提出并完善了基于案例的工业设计知识表达和基于情境的产品造型设计技术，建立了基于任务情境的数控机床造型复合信息模型，基于Client/Server架构建立了面向机床行业用户的造型设计支持软件平台。论文的研究过程从机床造型设计人员的需求和行为出发，对数控机床的设计过程、设计特征、设计心理模型等进行分析，结合设计调研、任务实验和意象认知实验等一系列实证方法，获取数控机床产品造型设计的领域本体知识。通过案例情境的知识表达新方式，对造型设计过程和经验进行封转和表达，实现了基于任务情境的复合信息模型，并以此为基础构建了包含300个造型案例，2000个造型方案的案例库系统。最后在反复验证与多个原型系统开发试验的基础上，建立了一个基于案例的计算机辅助产品造型设计系统——KB-ICAID，使得该系统具有与实际设计过程十分相似的项目管理、案例检索和再设计等实用功能并能够运用到实际开发工作中。通过在在国内几个大型机床主机厂进行示范应用，完成了若干款数控机床产品造型设计工作，验证了整个知识研究和系统构建的可行性和实用性。工业设计问题求解过程的复杂性和模糊性决定了现有的关于良性结构问题的知识获取、表达技术不能直接使用在工业设计领域中，需要专门的、针对性的知识研究方法和技术。论文研究借鉴了人工智能、心理学、社会学等研究成果，采用实证主义研究思想，提出了一套以心理学实验和调研为基础的、定性和定量相结合的设计知识获取、表达和建模方法，实现了对造型设计这种非良性结构的求解问题中的内隐性、经验性和形象化知识进行挖掘和提取，并将其作为研究产品造型设计过程和领域本体知识的获取和表达形式，最终提出了造型设计知识研究和情境案例表达的理论和复合信息模型。在以案例情境为核心的知识获取表达技术研究基础上，论文按照“以用户为中心”的软件开发思想先后构建了不同层次的基于知识的产品造型设计系统——ICAID、CBID和KB-ICAID等，实现了造型知识的案例化表达和任务情境构建，从而有效辅助机床设计人员进行数控机床的产品造型设计工作。系统经过评估验证后已经在多家机床企业进行了应用，协助企业开发了多款新产品并投产。更多还原

【Abstract】 Knowledge-based design has been the focus of modern design research andtransformation trend from traditional experience-based design. The progressive formdesign requirements of CNC machine tools, as the representative of complex products,has brought new challenges and opportunities to the traditional industrial designresearch and practices into new challenging&innovative topics. This paper initiates aknowledge-based assisted industrial design system concept by deploying a series ofstudies to facilitate it into realization. The thesis aims at the development ofknowledge-based design support system for CNC machine tools in order to solveintegrated form design problems of this typical complex product and to meet theinformation technology and design capability updating requirement for heavyequipment industry. The main content of this dissertation is the industrial designknowledge acquisition and expression issue of complex product at the system leveland subsequent implementation techniques, and the construction of suchknowledge-based Industrial Design assisted system. Being supported by several statelevel key research project, this thesis proposes and improves the case-based industrialdesign knowledge representation mechanism and scenarios based industrial designproduct design technology, then establish a task-scenarios based CNC form designintegrated information model, a user-oriented design support platform based on Client/Server architecture for industry.Based on the CNC machine tool design process, research started from thepersonnel needs and behaviors of designers with which design thinking (mental)models, process and habit were analyzed by carrying out series of empirical researchmethods including design survey, design task study and image congnition experiments,in order to excavate and obtain CNC machine product design domain ontologyknowledge. Through transfering and expressing of traditional experience knowledgeand process knowledge, research plan to achive a task-based situational informationmodel by case&scenario theory and then construct a design-case library systemwhich contains more than300s product form design cases and2000s design proposals.After repeated verification tests for subsequtial prototype system, the research finallydeveloped and established a knowledge case-based complex product form designassisted system: KB-ICAID, which being capable of stimulating the design process ofactual design project management, case retrieval, re-engineering and other practical functions and be able to apply on practical design projects. As the verification andapplication process, the KB-ICAID system has been practiced in some real designbusiness program with several large domestic machine tools company, illustrated thefeasibility and practicality of this knowledge supporting system and its construction.The complexity and ambiguity of industrial design problem solving process tellus the existing concerning benign knowledge acquisition, representation techniquesfor structural problems cannot be used directly in the field of industrial design. Thereis an imperative need for specialized and targeted knowledge research method andtechnique. Thesis draws on artificial intelligence, psychology, sociology and otherresearch results by following the imperical research ideology, sets up psychologyexperiments based, qualitative and quantitative combination method for designknowledge acquisition, expression and modeling. Tacit and experiential knowledge ofindustrial design process and experts’ knowledge were mined and then extracted intodesign knowledge system in order to support product design process and the domainontology knowledge acquisition. The main contribution of this thesis is the designknowledge presented case studies and situational theory and complex expressioninformation model.With the "user-centric" software development ideas, research have constructeddifferent levels of knowledge-based product design system-ICAID, CBID andKB-ICAID etc. to modeling design knowledge and design process by case-expressionand task scenarios conversion which can effectively assist designers in CNC machinetool industrial design field, the usability of such concept supporting system has alsobeen evaluated and applied in numbers of machine tool enterprises.更多还原