面向个性化车身产品客户动态需求获取与预测的扩展QFD研究  

An Extended QFD for Capturing and Forecasting Dynamic Customer Needs in Personalized Auto-body Development

作　　者：潘振华[1] 刘海江[1] 

机构地区：[1]同济大学机械工程学院,上海201804

出　　处：《机械科学与技术》2014年第4期564-572,共9页Mechanical Science and Technology for Aerospace Engineering

基　　金：国家高技术发展计划(863计划)项目(2008AA04Z105)资助

摘　　要：提出基于成分数据时间序列与向量自回归模型的扩展质量功能展开方法,在车身产品开发的初期,捕捉并预测客户的个性化需求。基于层次分析方法,建立传统的质量功能展开模型;根据客户需求重要度取样所形成的时间序列特点,将单纯形空间内的成分数据转换为实域内的时间序列,并对其应用向量自回归模型进行预测;将有效的样本外预测序列转换为客户需求重要度预测时间序列,计算出设计属性的预测优先度,以指导个性化车身产品开发过程中的资源分配。A novel extended quality function deployment （QFD） method based on compositional data time series and vector autoregression model was proposed to capture and forecast customers＂ individualized needs during the early stage of auto-body product development. Firstly, traditional QFD model was formulated in terms of analytic hierarchy process （AHP） technique. Secondly, according to the sampled customer importance ratings, the compositional data time series in simplex space was transformed into those in real space, which were forecasted by vector autoregression （VAR） model. Lastly, the validated out-of-sample forecasted time series was transformed back to compositional time series of customers＇ importance ratings in simplex space, and the priorities of design attributes were calculated se- quentially for guiding and planning the resources allocation in personalized Auto-body development.

关 键 词：质量功能展开 个性化车身 客户动态需求 成分数据 向量自回归模型 

分 类 号：O213.1[理学—概率论与数理统计] U462.1[理学—数学]