摘要
目的针对市场上产品外观设计缺少考虑用户感性需求的问题,采用在线评论与感性工学理论相结合的方法对产品外观进行情感设计研究,满足用户的情感需求。方法以鼠标为例,基于在线评论数据提取鼠标的感性意象和设计特征,结合产品解构的方法确定鼠标外观的设计要素,建立鼠标感性意象与设计要素之间的径向基函数神经网络模型,并采用粒子群优化算法对4个感性意象进行多目标优化,最后通过主观感性评价实验进行了验证。结果优化后的鼠标原型能够获得更高的感性意象得分,符合用户的情感需求。结论在线评论应用到感性设计的研究领域,极大地提高了用户调查的精确性和情感化设计的效率,为感性设计的相关研究提供了重要参考。
Objective Based on the problem that mouse design in the market lacks consideration of users'emotional needs,the method of combining online review and kansei engineering was used to conduct emotional design on the appearance of products to meet users'emotional needs.Methods The mouse was taken as an example.Kansei image and design elements of the mouse were extracted from online review data.Combined with the method of product deconstruction,concrete design elements of the mouse were determined.RBF neural network relationship model between the mouse's kansei image and design elements was established,and then the particle swarm optimization algorithm was used to perform multi-objective optimization on four kansei image words.Finally,the optimized result was verified by subjective kansei evaluation experiment.Results The results showed that the optimized prototype can obtain a higher kansei image score and meet users'emotional needs.Conclusion The application of online review to the research field of kansei design greatly improves accuracy of user surveys and efficiency of emotional design,which provides important implications for related studies on kansei design.
作者
陈佳豪
郭伏
任增根
张泽宇
姜钧译
CHEN Jia-hao;GUO Fu;REN Zeng-gen;ZHANG Ze-yu;JIANG Jun-yi(Northeastern University,School of Business Administration,Shenyang 110169,China;National Natural Science Foundation of China,Center for Science Communication and Achievement Transformation,Beijing 100085,China)
出处
《人类工效学》
2021年第4期6-12,共7页
Chinese Journal of Ergonomics
基金
国家自然科学基金项目(71771045、72071035)。
关键词
工业设计
可用性
正交试验
在线评论
感性评价
情感化设计
多目标优化
RBF神经网络模型
industrial design
usability
orthogonal experinent
online review
kansei evaluation
emotional design
multi-objective optimization
radial basis function neural network model