Construction of the Expert System for the Selection of Materials using the Kansei Engineering

设计符合用户品味的产品是很重要的，因此，设计师应该了解材料特性和心理特性(感性特性)之间的关系。本研究以办公椅质感面料为例，探讨了视觉形象与纺织面料结构元素之间的关系。首先对纺织品样品进行感官评价，建立织物样品的视觉形象。在感官评价中，学生对34种纺织面料进行了20个评分表的评价。然后使用因子分析对这些数据进行分析。以织物的组织结构、厚度、密色比为参数，结合纺织设计师的建议，建立了织物视觉形象与组织结构、厚度、密色比之间的神经网络模型。将这些参数用于输入，然后对视觉图像进行预测。研究的结果表明，对于具有周期性图案的织物，可以通过使用这些参数来改变视觉图像。

It is important to design products to suit the users' tastes, therefore, designers should understand the relation between material properties and psychological characteristics (kansei properties). In this study, the relation between visual image and the structural elements of textile fabrics was discussed as a case of texitile fabrics for office chairs. At first, a sensory evaluation was made to establish the visual image of textile fabric samples. Then the relation between them was investigated by the use of a neural network.In the sensory evaluation, 34kinds of textile fabrics were evaluated by students on 20 rating scales. This data was then analyzed using factor analysis. Six factors were obtained as follows ; familiarity, dignity, smoothness, uniformity, richness, and dampness.Four elements of textile fabrics were chosen as the parameters ; weave organization, thickness, ratio of density and color, according to the standards of textile fabrics and advice from textile designers.Finally, a neural network model was constructed to investigate the relation between the parameters and the visual image. The parameters were applied to input, then the visual image was predicted. As a result of the investigation, it became clear that the visual image can be changed by the use of the parameters in case of the textile fabrics with periodic patterns.