(2016) used the two variant autoencoders, the contractive autoencoder (CAE) and the denoising autoencoder (DAE) to develop a nonlinear process monitoring model. However, these methods are inherently limited in that they are not capable of taking in nonlinearity, and various data processing methods have been implemented to resolve this problem. Among the various methods, conventional data-driven monitoring methods, such as PCA, FDA, and ICA have been used and improved extensively. Process monitoring is an essential part of safe plant operation, and has been studied in various angles over the past few decades. Won Bo Lee, in Computer Aided Chemical Engineering, 2018 1 Introduction 13th International Symposium on Process Systems Engineering (PSE 2018)Ĭhangsoo Kim.
0 Comments
Leave a Reply. |