Output-relevant Variational autoencoder for Just-in-time soft sensor modeling with missing data. (August 2020)

• Record Type: Journal Article
• Authors: Guo, Fan; Bai, Wentao; Huang, Biao
• Is Part Of: Journal of process control. Volume 92(2020)
• Page Start: 90
• Page End: 97
• Publication Date: 2020-08
• Subjects: Output-relevant Variational Autoencoder -- Just-in-time modeling -- Symmetric Kullback–Leibler divergence -- Gaussian process regression -- Missing observations; Process control -- Periodicals; Fabrication -- Contrôle -- Périodiques; Process control; Periodicals; Electronic journals; 660.281
• DOI: 10.1016/j.jprocont.2020.05.012
• Languages: English
• ISSNs: 0959-1524

Abstract

Abstract: Main challenges for developing data-based models lie in the existence of high-dimensional and possibly missing observations that exist in stored data from industry process. Variational autoencoder (VAE) as one of the deep learning methods has been applied for extracting useful information or features from high-dimensional dataset. Considering that existing VAE is unsupervised, an output-relevant VAE is proposed for extracting output-relevant features in this work. By using correlation between process variables, different weight is correspondingly assigned to each input variable. With symmetric Kullback–Leibler (SKL) divergence, the similarity is evaluated between the stored samples and a query sample. According to the values of the SKL divergence, data relevant for modeling are selected. Subsequently, Gaussian process regression (GPR) is utilized to establish a model between the input and the corresponding output at the query sample. In addition, owing to the common existence of missing data in output data set, the parameters and missing data in the GPR are estimated simultaneously. A practical debutanizer industrial process is utilized to illustrate the effectiveness of the proposed method. Highlights: An output-relevant VAE is proposed for extracting output-relevant features by using correlation between process variables. With symmetric Kullback–Leibler divergence, the similarity is evaluated between the stored samples and a query sample. According to the values of the SKL divergence, data relevant for modeling are selected. Owing to the common existence of missing data in output data set, the parameters and missing data in the GPR are estimated simultaneously. A practical debutanizer industrial process is utilized to illustrate the effectiveness of the proposed method.