基于主成分分析与TOPSIS模型相结合的函数型产品质量特性的优化方法研究  

Optimization of Functional Product Quality Characteristic Problems Based on PAC and TOPSIS Model

作　　者：许静[1] 何桢[1] 袁荣[1] 陈喆芝[1] 

机构地区：[1]天津大学管理与经济学部,天津300072

出　　处：《工业工程与管理》2016年第3期59-67,共9页Industrial Engineering and Management

基　　金：国家自然科学基金杰出青年基金资助项目(71225006)

摘　　要：基于稳健参数设计理论提出了一种将主成分分析法与逼近理想点决策方法(TOPSIS)相结合的非线性轮廓图(Non-Linear profile)优化方法。首先利用两步建模法拟合响应模型,计算模型参数的满意度函数值,其次对模型参数的满意度函数值进行主成分分析,消除参数之间的相关性,并构建模型参数变异模式图,确定选定主成分的优化方向。最后利用TOPSIS模型求得选定主成分的最优贴近度(OPI),将其作为最终的优化指标。传统的优化方法都忽略了模型参数之间的相关性及优化过程的稳健性,并且需要复杂的数学计算,而本文所提方法可以有效解决这些问题。最后利用该方法对文献中的实例进行了分析研究,证明本文方法切实可行,优化结果令人满意。A method to optimize Non-Linear profile based on the TOPSIS model and PCA is presented.Firstly,the method of two-step modeling fitting response model is adopted,and desirability function is used to measure the model parameters.Principal component analysis is then performed on the desirability function values to obtain a set of uncorrelated components.The optimization direction for each component is also determined based on the corresponding variation mode chart.Finally,technique for order preference by similarity to ideal solution(TOPSIS)is adopted to derive the overall performance index(OPI).The traditional optimization method ignores the relationship between the parameters and the robustness of the optimizing process.The effectiveness of the proposed method is confirmed through a numerical example to solve these problems.

关 键 词：函数响应 TOPSIS 主成分分析 参数设计 

分 类 号：F406.3[经济管理—产业经济]