基于主成分分析的产品外观设计决策模型  

Decision-making Model of Product Design Based on Principal Component Analysis

作　　者：杨程[1] 孙守迁[1] 刘征[1] 柴春雷[1] 

机构地区：[1]浙江大学,杭州310015

出　　处：《中国机械工程》2011年第18期2218-2223,共6页China Mechanical Engineering

基　　金：国家自然科学基金资助项目(60975048;51005203)

摘　　要：利用主成分分析对产品外观设计方案进行多目标决策,回避了设计评价中常用的评价指标模糊权重系数求解,增强了设计决策的客观性。首先,根据设计师访谈、用户调研和国内外的设计评价研究泛选产品设计评价指标;通过评价指标的公信度评测、重合度评测和效用度评测,消减重复、低效的设计评价指标,确定产品设计评价指标体系;基于主客观指标赋值,建立评价矩阵;最后,计算评价矩阵的相关系数矩阵,求解其主成分及其贡献率,构建综合评价函数。通过具体设计方案评价实验,各主成分代表的意义符合工业设计师的认知,最终评价结果与市场和用户综合评价基本一致。该研究在一定程度上增强了设计评价的客观性和科学性,对产品外观设计决策有实际价值。To reduce the subjective factors in the decision--making process and the dimension of the evaluating indicators, a methods based on principal component analysis was proposed in the corn prehensive evaluation of the design plan. First, based on correlational research,some product design evaluating indicators were selected primarily. Then, the system of product evaluating indicators was set up by credibility analysis, coincidence analysis and utility analysis. Based on group evaluation, the part of subjective indicators was assigned. All values of evaluating indicators were standardized. Finally, the main components of correlation matrix which was built by evaluating indicators were solved to build com- prehensive evaluation function. The meanings of the principal components met the designer＇s cognition, the final evaluation results were consistent with market and users evaluation. To some extent, the study makes evaluation of product design more objective and scientific, and has a real value.

关 键 词：多目标决策 主成分分析 群体评价 设计评价 

分 类 号：TP391.7[自动化与计算机技术—计算机应用技术]