PROJECT TITLE :

Supervised Latent Factor Analysis for Process Data Regression Modeling and Soft Sensor Application

ABSTRACT:

This transient proposed a new supervised latent factor analysis (FA) technique for method knowledge regression modeling. Totally different from the ancient principal part analysis/regression model, the new model will successfully estimate heterogeneous variances from totally different method variables, which is a lot of sensible. Underneath the identical probabilistic modeling framework, the single supervised latent FA model is further extended to the mixture form. Efficient expectation–maximization algorithms are developed for parameter learning in both single and mixture supervised latent FA models. Based on the regression modeling between simple-to-measure and difficult-to-live method variables, two soft sensors are engineered for quality prediction in the process. 2 case studies are provided to evaluate the modeling and performances of the new strategies.