【摘要】 目的运用感性工学和BP神经网络构建产品质感调和设计模型。方法通过多维尺度分析和聚类分析,分析出了最适合于描述质感调和的代表性感性意象词汇集;通过统计分析,确定了以透明度、光泽度和粗糙度及其细分类目为元素的质感调和空间;通过质感调和实验,获得了针对实验样本的质感空间特征值和感性评价值;最后通过MATLAB的BP神经网络工具箱,建立了基于感性意象的质感调和设计BP模型,并重新设计检验样本论证了模型的可靠性。结论该模型为后续的计算机辅助质感调和设计系统提供了理论和算法依据,可以提高产品设计的成功率。更多还原

【Abstract】 A product texture harmony design model was built based on the kansei engineering and the BP neural network. The representative kansei words that were most suitable to describe the texture were selected by multidimensional scaling analysis and cluster analysis. Through statistical analysis, texture harmony space was determined with degree of transparency, gloss, roughness and other disaggregated classifications as elements. Texture space characteristic value and kansei appraisal value of sample were obtained through texture harmony experiment. Finally the texture harmony design BP model was worked out based on the kansei image via BP neural network tools box of MATLAB, and then the sample was redesigned and verified to demonstrate the reliability of the model. The model provides the theory and algorithm basis for the subsequent computer-assisted texture harmony design system, and also improves the success rate of product design.更多还原