Multistage Monitoring of Batch Processes Using PLS
Multiway principal component analysis (MPCA) has been extensively applied to the batch process monitoring. In the case of monitoring a two-stage batch process, the inter-stage variation is neglected, if MPCA models for each individual stage are used. On the other hand, combine two stage reference data into a large dataset that MPCA is applied to, the dimensions of unfolded matrix will increase dramatically. Furthermore, in the on-line monitoring phase, the condition of an operation can only be determined after two-stage batch runs were completed. In this paper, partial least squares (PLS) is applied to monitor the inter-stage relation of a two-stage batch process. In post-analysis of abnormalities, it can clarify whether root causes are from the previous stage operation or due to the changes of the inter-stage correlations. This approach is successfully applied to a semiconductor manufacturing process.
Batch Process monitoring principal component analysis consensus PCA hierarchical PCA,canonical variate analysis partial least squares
Jialin Liu David Shan Hill Wong
Department of Information Management,Fortune Institute of Technology,1-10,Nwongchang Rd.,Neighborhoo Department of Chemical Engineering,Tsing Hua University 101 Section 2,Guang Fu Road,Hsinchu.
国际会议
西安
英文
2007-08-15（万方平台首次上网日期，不代表论文的发表时间）