【作者】 鲁虹；

【导师】 许星； 陈雁；

【作者基本信息】 苏州大学 ， 服装设计与工程， 2010， 博士

【摘要】 本课题针对服装的感性特点,对感性设计方法进行了系统的研究。总结并分析了服装设计的各项要素、服装的感性特征、二者的关联以及服装设计规则,构建了服装感性设计的知识平台。提出了进行专家知识获取与形式化表征、基于知识平台进行服装交互设计、在推荐交互过程中实现专家认知与客户认知统一等研究的完整思路。研究以男西装为实例,对知识平台的构建过程、方法和结果进行了分析。在该知识平台上,建立了可用于交互设计的应用系统。研究成果不仅可以用于男西装设计,亦可推广运用至其他服装。对服装设计个性化、感性化、智能化、快速化等目标的实现有着重要的指导意义。本研究对服装感性设计知识的构成进行了必要的界定。根据服装设计流程,将服装感性设计知识划分为:服装设计要素、服装感性特征、服装设计要素与感性特征的映射以及服装设计规则四个部分。同时,对各部分知识的特点、涵盖内容进行了分析。这部分研究为服装感性设计知识的表达、知识平台的构建提供了基础。本研究还针对服装感性设计知识获取与形式化表征的技术路线与方法进行了分析。采用访谈法、阶梯法对服装设计要素进行了分解,并采用层次结构法加以表达,构建了服装设计要素空间。以男西装为研究对象,将款式要素细分为9个要素类目(轮廓造型、刻口形态、串口位置、门襟开度、驳头宽窄、门襟外缘、胸袋设计、腰袋设计、装饰设计)、28个要素子类,组合后形成26244种款式;将色彩要素进行细分,不同的色相、明度、纯度,组合后形成89种色彩,其中有彩色79种、无彩色10种,用于男西装设计推荐及后续研究;将面料要素从表观特征与内在特征两部分进行细分。通过访谈法、卡片法、层次分析法与模糊理论,建立了相应的感性意象空间,并对专家在感性意象与设计要素类目的映射规则、设计要素项目权重等知识进行获取与表示。引入三角模糊数对七种语意描述(非常、很、比较、一般、比较、很、非常)进行量化表示。采用访谈法,资料查阅法等对服装设计中的程式化规?则,基于客户体型的服装设计规则、基于客户肤色的服装设计规则进行获取,并以表格形式加以表示。研究以男西装为例,在广泛收集、初步筛选基础上,对描述男西装感性的68个形容词进行语意亲密程度归类,形成具有4对形容词组的感性意象空间(正式的——休闲的,儒雅的——粗犷的,华贵的——朴素的,古典的——现代的),用以描述男西装的感性意象。同时对男西装设计中的相关设计规则进行了有效的获取和形式化表示。在上述研究的基础上,对基于客户需求的模糊推荐与基于客户认知的服装交互设计进程进行研究。结合客户基本信息,以知识平台为依托,进行基于知识的推荐。通过客户感性意象需求,采用层次分析法得到的客户多需求权重,引入模糊综合评判理论,建立满足条件的服装集,进行基于单一需求与多重需求的模糊推荐。推荐过程中,根据客户评价,调整所推荐的服装集,进行交互设计,直至得到客户满意的结果。根据交互结果的海量数据,建立客户知识获取与挖掘模型。基于联合分析法,选取款式样本确定实验方案与数据处理方法,获取客户在设计要素类目、要素子类与感性意象映射的知识,并与专家知识对比,为专家认知与客户认知的统一,知识平台的更新、完善提供必要的参考。最后,将计算机技术、人工智能技术引入服装设计过程,进行了服装感性设计应用系统的设计与实现。系统前台采用Visual Basic语言编写,后台数据库采用Access,前台与后台数据的连接通过VB提供的ADO接口来实现。整个系统主要由数据库、界面与三大模块(知识管理模块、服装推荐与交互模块、客户知识获取模块)构成。实现了知识管理、数据存放、服装推荐与交互、客户知识获取等基本功能。更多还原

【Abstract】 In this paper, the methods for clothing sensory design were studied systematically according to its characters. The clothing design elements, semantic expression, reflection between the former two items, and the design rules were discussed, analyzed and represented. Then, the knowledge base for clothing sensory design was constructed. The whole research idea of the acquisition and representation of the experts’knowledge was put forward. On the basis of knowledge acquired, the procedure of clothing recommendation and interactive design was introduced and analyzed which could improve the cognition consistency between the designers and the customers. The men’s suit was used as a case study to explain the whole procedure, methods and results of knowledge base construction. All techniques and methods were not limited to men’s suit itself but extended to other garments or products. The conclusions of the paper would play an important role in realizing the goals such as clothing individuation, sensation, intellectualization and fashion etc..The content of the knowledge for clothing sensory design was defined. It involved four interrelated parts, namely clothing design elements, the clothing semantic expressing, reflection between the design elements and the semantic expression, and the clothing design rule. The character and contents of each part were discussed and analyzed respectively. The results would be the base of the further knowledge acquisition and representation and moreover the construction of the knowledge base.The technique path and method of experts’knowledge acquisition and representation were analyzed. The design elements were discussed, analysed and decomposed by using interviewing and laddering methods and then represented in the form of hierarchy structure so as to build the design elements space. The men’s suit was used to explain whole procedure and got some results. The style design elements of the men’s suit were decomposed into 9 items: silhouette, style line of collar, fall, break point, rever position, facing corner, bust pocket, waist pocket and decoration. And the items could be decomposed further into 28 categories. 26244 kinds of clothing style could be got by the elements combination. The elements of clothing color design were discussed and analyzed, and then the 89-kind-color sample base (79-kind-chromatic color and 10-kind achromatic color) was obtained for further recommendation on the base of three design elements hue, value and chroma. The elements of clothing material design was discussed and analyzed from both extrinsic character and intrinsic character.The interviewing, card sorting, AHP (analytic hierarchy process) and fuzzy mathematics theory were used not only to build the corresponding semantic space but also to acquisition the reflection rule between the design elements space and semantic space. The triangular fuzzy numbers were introduced to quantify the 7-point scales of linguistic descriptions (extreme, very, rather, average, rather, very and extreme). The design rule of the men’s suit was obtained by the method of literature reviewing and interviewing. It included both the basic rule derived from the beauty rule and the design principle based on consumer’s character such as body type and facial color etc. The corresponding knowledge of the design rule was obtained, collated and recorded in the form of table. During the research of the men’s suits, a large amount of adjective words related to the sensory were collected. After the primary analysis discussion and filtration, 68 adjective words were prepared for card sorting. According to the intimate degrees between the words, the semantic space of the sensory evaluation towards the men’s suit was modeled in the form of 4 word-pairs: formal- casual, classical- modern, gorgeous- simple and elegant- masculine Moreover, the relative design rules of men’s suit were acquired and represented.On the basis of the research above, the process of customer-based fuzzy recommdation and interactive design were studied. Based on the knowledge base and consumer’desirability acquired, the knowledge-based recommendation was realized. The customer needs was ranked on basis of their importance by using the technique of AHP. Based on the theory of the fuzzy synthesis evaluation, the customers’requirement was analysed, and all samples which was composed of different design elements could be sorted into different sets respectively. The recommendation results towards the sole or multi requirements were the clothing set including lots of clothing from which the consumers can choose their preferences. During the process of the fuzzy recommendation, the clothing set can be adjusted and the interactive design could be done on the basis of the appraisal of the customers until the satisfactory result appeared. Through the data mining, the customers’knowledge acquisition and mining model would be built and the cognition of the customers would be understood. CA (the conjoint analysis) method could be used and the proper profiles including multiple conjoined design elements were chosen and the customers’knowledge of the relation between the design elements space and semantic space was acquired. The acquisition of customer’s knowledge could improve cognition consistency between the experts and customers and provide reference to the renewal and perfection of the knowledge base.Finally, based on the research conclusions above, the computer technology and artificial intelligence technology were involved and the application system for clothing was designed and developed. The system was developed by VB programming language, database management software Access and ADO connection, which was composed of database, interface and three modules (knowledge management module, clothing recommendation and interactive design module and customers’knowledge acquisition module). The function such as knowledge management, data memory, clothing recommendation, interactive design and customers’knowledge acquisition etc. could be realized in this system.更多还原