%0 Journal Article
%T 支持向量机回归模型在航空座椅造型眼动评价中的应用<br>Application of Support Vector Machine Regression Model in Aircraft Seating Styling Eye Movement Evaluation
%A 李文华
%A 余隋怀
%A 于明玖
%A 田保珍
%J 机械科学与技术
%D 2018
%X 为解决造型评价中主观评价方法缺少定量数据支持的问题，提出一种基于眼动参数的造型评价方法。以航空座椅为研究对象，利用SMI公司的RED遥测型桌面式眼动追踪系统，采集了14名被试在对8种航空座椅造型进行评价时的眼动数据，并获取了被试对航空座椅外观造型的主观评价值，建立了以多项眼动参数为输入的基于支持向量机回归的航空座椅外观造型评价模型，对模型进行了验证，并与其它方法进行了对比。结果表明，支持向量机回归模型的评价结果优于人工神经网络模型，具有较高的准确性。研究结果有助于航空座椅设计企业提高座椅造型评价的效率，降低成本并缩短开发时间，同时，也有助于航空公司的座椅选型工作。本文利用眼动参数建立的造型评价模型结准确率较高，适合进一步推广应用，为产品的造型评价提供了一种更为高效的评价方法。<br>To address the disadvantage that subjective evaluation method lacks quantitative data support in styling assessment process, a styling evaluation method based on eye movement data was developed. This paper took the airplane seat as the research object, used the RED remote desktop type eye tracker of SMI company to collect 14 subjects' eye movement data during the evaluation process of 8 different airplane seats, and obtained the subjective ratings to eight airplane seats. A mathematical model about airplane seat styling evaluation based on support vector machine, which used five eye movement parameters as inputs, was established and verified. The result showed that it had higher accuracy, compared with neural network model. The research helps enterprise improve efficiency of airplane seats design and evaluation, and reduce cost and development time; it also benefits the airliners' seating type selection. In consideration of good accuracy of styling evaluation model, it is suitable for further popularization and application; it provides an efficient evaluation method for the product design evaluation
%K 航空座椅
%K 外观造型
%K 眼动
%K 评价方法
%K 支持向量机回归&lt
%K br&gt
%K aircraft seating
%K styling
%K eye movement
%K evaluation method
%K support vector machine regression
%U http://journals.nwpu.edu.cn/jxkxyjs/CN/abstract/abstract7175.shtml