摘要
针对国内经济型轿车的设计、质量不好和日益增长的需求之间的矛盾。必须从客户角度设计,控制汽车制造质量。为了解决客户主观、模糊化的需求信息,必须把质量需求信息定量化。结合人对质量认知的绝对与相对因素,必须从企业形象、使用质量和性价比三个因素进行客户质量需求分析。根据人的认知过程的特殊性,三个因素之间存在很大的相关性。三个因素分别根据层次法建立指标体系,在假设三个因素存在线形相关的前提下,建立结构方程组模型。为了避免人为构建的指标之间的共线性关系,以偏最小二乘法为基础进行数据处理,以追求更真实,科学的分析结果。通过客户需求信息的定量化,将可以形成以数据驱动的形式指导企业的设计和质量生产行为。
In order to solve the conflictton of bad quality aria increasmg market, it＇s strongly necessary to control quality from the aspect of customers. The challenge is to translate customers needs infarma- tion from qualitative and fuzzy, to quantitative. Because of human cognitive process to evaluate quality, the evaluation model must include factors： use quality, performance-price ratio and brand. The three factors are strongly interactional and to establish evaluation items by AHP method. And supposing three factors just have linear relation to establish Structural Equation Model. The Partial Least-Squares Regression arithmetic can avoid interaction of different items, and the result will be scientific and objective. By the process of translating customer information, it will instruct manufacturer to better designing and purposeful to control quality.
出处
《机械设计与制造》
北大核心
2007年第7期184-186,共3页
Machinery Design & Manufacture
关键词
质量评价
偏最小二乘回归
结构方程组
经济型轿车
Quality evaluation
Partial least-squares regression
Structural equation model
Economic automobile
