针对制造过程中的多元工序质量控制问题,提出了一种基于联合控制域变换的工序质量控制算法。利用主成分分析法,将原工序质量变换为各分量之间彼此无关的一组新工序质量,使得新工序质量的联合控制域为多维空间中各半轴与对应坐标轴重合的超椭球体,并包含于由新分量控制界限交集构成的超长方体之内,从而将工序质量控制体系中所需的控制图数目减少为1,实现对控制体系的简化。理论分析和实验表明,基于联合控制域变换的多元工序质量控制体系能够对工序质量的异常进行有效监控,提高控制体系的运行效率。
Aiming at the problem of multivariate process quality control in manufacturing,an algo-rithm based on j oint control domain transformation is proposed.Original quality vector is trans-formed to a new one whose components are unrelated by using the principal components analysis method,thus the joint control domain of new quality vector is a hyper ellipsoid whose axes are co-incided with the corresponding axes in multidimensional space,and contained in the hyper cube formed by the intersection of every component control limits,so the structure of process quality control system is simplified and the number of control charts is reduced to 1 .Theoretical analysis and experiments prove that the proposed quality control system based on j oint control domain transformation can monitor the abnormal process quality effectively,and the efficiency of process quality control system is improved.