基于情感分析的汽车造型设计感性评价方法  被引量:6

Perceptual Evaluation Method of Automobile Styling Design Based on Sentiment Analysis

在线阅读下载全文

作  者:孔德洋 黄偲蕊 麻殊捷 KONG Deyang;HUANG Sirui;MA Shujie(School of Automotive Studies,Tongji University,Shanghai 201804,China)

机构地区:[1]同济大学汽车学院,上海201804

出  处:《同济大学学报(自然科学版)》2022年第12期1817-1824,共8页Journal of Tongji University:Natural Science

基  金:上海市科委科普项目(20DZ2306500)。

摘  要:针对分析用户感性评价时出现的数据采集范围不够广、成本高、数据准确性低、非结构化数据利用率低等问题,提出一种基于情感分析的工业设计感性评价方法,以纯电动汽车为例,构建电动汽车造型感性意象的预测模型。首先基于情感分析量化用户的主观评论,获得感性意象评分,然后针对纯电动汽车的造型特征,提取了与传统内燃机汽车存在差异的主要设计要素,最后通过数量化理论I建立感性意象评分与设计要素的映射模型。以电动汽车设计方案优化为例,证明了该方法的可行性,预测结果的相对误差为5.41%。该方法符合用户的主观认知,能够有效预测用户对方案的感性评价,从而辅助设计师根据感性评价优化产品设计,具有实际应用价值。Aiming at the problems of insufficient data collection range,high cost,low data accuracy and low utilization rate of unstructured data when analyzing user perceptual evaluation,a perceptual evaluation method of industrial design based on sentiment analysis was proposed.Taking battery electric vehicle as an example,the prediction model of perceptual image of electric vehicle modeling was constructed.Firstly,the subjective comments of users are quantified based on sentiment analysis to obtain the perceptual image score.Then,according to the modeling characteristics of battery electric vehicles,the main design elements different from traditional internal combustion engine vehicles are extracted.Finally,the mapping model between perceptual image score and design elements is established through quantitative-I theory.Taking the optimization of electric vehicle design scheme as an example,the feasibility of this method is proved,and the relative error of the prediction result is 5.41%.This method conforms to the user’s subjective cognition and can effectively predict the user’s perceptual evaluation of the scheme,so as to assist the designer to optimize the product design according to the perceptual evaluation.It has practical application value.

关 键 词:工业设计 汽车造型 情感分析 数量化理论I 感性工学 电动汽车 

分 类 号:TB472[一般工业技术—工业设计] TP391.1[自动化与计算机技术—计算机应用技术]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象