【摘要】 针对PCA单尺度建模的局限性,综合应用小波分析和主元分析PCA,将小波分析提取变量局部特征的能力及主元分析分解变量线性关系的优势结合起来,引进新的多元统计过程监测工具MSPCA,并将其应用于温度周期化的流向变换催化燃烧过程。研究表明,与PCA相比利用MSPCA可以更有效地监测到过程的异常状况。更多还原

【Abstract】 To improve the ability of PCA to monitor the statistical process,a new kind of monitoring tools named Multiscale Principal Component Analysis(MSPCA) is introduced which combines the ability of PCA to decorrelate the variables by extracting a linear relationship,with that of wavelet analysis to extract deterministic features and approximately decorrelate autocorrelated measure-ments.Compared with PCA,MSPCA is more effective for process monitoring.The superior performance of MSPCA is illustrated by applying it to reversed-flow catalytic combustion process monitoring.更多还原