基于信息熵的概念设计的质量评价  

Quality evaluation of quality design based on entropy

作　　者：邓军[1,2] 余忠华[1] 吴昭同[1] 

机构地区：[1]浙江大学机械与能源工程学院,浙江杭州310027 [2]台州学院数学与信息工程学院,浙江临海317000

出　　处：《浙江大学学报（工学版）》2009年第8期1480-1484,共5页Journal of Zhejiang University：Engineering Science

基　　金：国家"863"高技术研究发展计划资助项目(2002AA413510);国家自然科学基金资助项目(50835008)

摘　　要：为避免不合理的概念设计给产品质量带来损失,建立了一般性的多层次的概念设计质量评价指标体系.线性规划了质量评价,分析了用户需求、顾客竞争性以及指标之间的互相关性等因素对质量评价指标权重具有的不确定性影响,说明了线性规划的权重的不确定性、线性规划的线性组合系数的不确定性、评价者对不同指标的评价的不确定性,并引入信息熵来描述这些不确定性.根据拉格朗日乘子法的法则,对质量评价的线性规划优化问题构造拉格朗日函数,并应用最大信息熵原理进行优化,求解概念设计质量评价的线性规划优化问题.该方法应用于汽车概念设计方案质量评价与优选,能够优选出最佳的概念设计方案,从而保证产品质量.To avoid the quality loss to the product caused by unreasonable product conceptual design, a generic multi-hierarchy quality evaluation system of conceptual design and a linear programming model of quality evaluation were set up. After analyzing the uncertain influences on the quality evaluation index weight by the importance of customer demand, the customer competition and the cross-correlation of evaluation indexes, the uncertainty of each evaluation index weight, the uncertainty of coefficients of linear combination in linear programming, and the uncertainty of valuator＇s evaluating for different evaluation indexes were illustrated. Thus, entropy was introduced to describe the uncertainties. According to the Lagrange law, a Lagrange function was constructed for the linear programming optimization of quality evaluation, and Jaynes maximum entropy principle was applied to the optimization solution of quality evaluation. In the last, this method was applied to the quality evaluation of car＇s conceptual designs for selecting the best design, and the results showed that the evaluation method was effective.

关 键 词：信息熵 概念设计 质量评价 不确定性 

分 类 号：TP202[自动化与计算机技术—检测技术与自动化装置]