• CN:11-2187/TH
  • ISSN:0577-6686

机械工程学报 ›› 2018, Vol. 54 ›› Issue (23): 126-136.doi: 10.3901/JME.2018.23.126

• 数字化设计与制造 • 上一篇    下一篇

基于脑电的产品意象推理模型研究

杨程1, 陈辰2, 唐智川3   

  1. 1. 浙江大学城市学院工业设计系 杭州 310015;
    2. 浙江大学现代工业设计研究所 杭州 310027;
    3. 浙江工业大学工业设计研究院 杭州 310014
  • 收稿日期:2017-11-22 修回日期:2018-05-09 出版日期:2018-12-05 发布日期:2018-12-05
  • 通讯作者: 杨程(通信作者),男,1978年生,博士,教授。主要研究方向为计算机辅助设计、设计知识、数字化设计。E-mail:yangchengyc@126.com
  • 作者简介:陈辰,女,1993年生,硕士研究生。主要研究方向为计算机辅助设计。E-mail:395249909@qq.com;唐智川,男,1987年生,博士,助理研究员。主要研究方向为脑机接口。E-mail:ztang@zjut.edu.cn
  • 基金资助:
    国家自然科学基金(61702454)和浙江省自然科学基金(LY18E050014)资助项目。

Study of Electroencephalography Cognitive Model of Product Image

YANG Cheng1, CHEN Chen2, TANG Zhichuan3   

  1. 1. Department of Industrial Design, Zhejiang University City College, Hangzhou 310015;
    2. Modern Industrial Design Institute, Zhejiang University, Hangzhou 310027;
    3 Industrial Design Research Institute, Zhejiang University of Technology, Hangzhou 310014
  • Received:2017-11-22 Revised:2018-05-09 Online:2018-12-05 Published:2018-12-05

摘要: 将用户对产品的感性认知通过脑电数据量化,探求消费者认知过程脑电与产品意象的对应关系,并以此建立相应的意象推理模型。以吊灯为研究样本,通过多组感性意象形容词描述产品意象空间。基于被试的行为数据、脑电信号和事件相关电位,分析被试在吊灯产品意象辨别过程中的决策行为。研究不同意象匹配情况下,各脑区N200、P300、N400等脑电成分的波幅和分布情况,确定脑电成分与产品意象匹配的映射关系。基于被试的反应时长、选择率和脑电信号等评价因素,构建模糊推理的产品意象认知模型。通过试验验证,推理模型与意象语义评估的结果有一致性,具有较好的可靠性。

关键词: 产品意象语义, 模糊逻辑, 脑电(EEG), 事件相关电(ERP), 意象推理模型

Abstract: Based on the EEG experiment, the user's perceptual judgment of the product is quantified. The corresponding relationship between the user's EEG information and the product image in the cognitive process is constructed, and the corresponding image reasoning model is established. The product image is described by multiple groups of perceptual image adjectives. Based on the behavior data, EEG and event related potential of the subjects, the decision behavior of the subjects during the image cognition is analyzed. With different images matching conditions, the amplitude and distribution of ERP components are studied to determine the mapping relationship between EEG information and product image. Based on the evaluation factors such as time length, selection rate and EEG signal, the product image reasoning model is constructed. It is proved that the results of the reasoning model and the image semantic evaluation are consistent and have good reliability.

Key words: EEG, event-related potential (ERP), fuzzy logic, image inference model, product image semantics

中图分类号: