复杂制造系统中波动源的探测和分离

造成产品质量缺陷的根本原因是波动 ,而减小制造系统波动的先决条件是确定其波动源。首先 ,本文根据制造系统中的相关矩阵构造了相应的赋权网络 ,依据网络和最大支撑树的聚类准则对制造系统中的各个过程进行分类。在此基础上 ,利用残差分析技术 ,确定并分离出制造系统的主要波动源 ,从而减少了系统波动 ,实施了稳健设计。

Detection and Isolation of the Variation Sources in a Complex Manufacturing System

This paper studies the detection of variation in the manufacturing system consisting of many processes with close correlativity. First, the paper introduces the background of reducing variation in the manufacturing system and establishes a weighted network corresponding to the correlation matrix of all processes, then it classifies all processes by clustering criterion of maximum spanning tree. On this basis, it determines the main variation source of each process in a class by residual analysis. The paper ends with the conclusions.;