摘要
针对控制过程中多变量的共线性,提出一种主成分质量控制方法,对化工中甲烷气体生产过程进行质量监控与诊断。首先运用在正常状态下所观测的13组样本数据建立过程控制模型,然后根据所建模型对在异常状态下观测的12组样本数据进行监控与诊断。结果表明:该方法能够及时地监测出故障并能准确查找出引起故障的过程变量。
To the deficiencies of commonly used multivariate ststistical process control, the principal component quality control method is put forward to proceed quality monitoring and diagnostics on the production process of practical industrial methane gas. First, the process control model was estab- lished using the first 13 groups of sample data we observed in the normal state. Then, we used the es- tablished model to proceed quality monitoring and diagnostics o~ the known 12 groups of sample data we observed in the abnormal state. The results show that the method can monitor the fault timelier and find out the process variables that cause malfunction accurately.
出处
《重庆理工大学学报（自然科学）》
CAS
2014年第1期96-101,共6页
Journal of Chongqing University of Technology：Natural Science
基金
重庆市科委资助项目(cstc2012jjA00018)
重庆市教委项目(KJ130810)
关键词
主成分分析
故障监控与诊断
T^2控制图
principal component analysis （PCA）
fault detection and diagnosis
controlling figureof T^2
