Industrial polyethylene melt index prediction using ensemble manifold learning–based local model

Abstract

For online melt index prediction in multiple‐grade polyethylene polymerization processes, using only a fixed model is insufficient. Additionally, without enough process knowledge, it is difficult to select suitable input variables to accurately construct prediction models. A novel manifold learning based local probabilistic modeling method named ensemble just‐in‐time Gaussian process regression (EJGPR) is developed. By utilizing output variables, an optimization framework is proposed to preserve the local structure of both input and output variables. Then the output information is integrated into construction of a JGPR‐based local model. Additionally, some new extracted variables in the projection space can be obtained. Moreover, using the probabilistic prediction information, the uncertainty of each JGPR‐based local candidate model can be simply described. Consequently, using an efficient ensemble strategy, a more accurate EJGPR prediction model can be constructed online. The melt index prediction results in an industrial polyethylene process show it has better performance than conventional methods. © 2017 Wiley Periodicals, Inc. J. Appl. Polym. Sci. 2017, 134, 45094.

Hybrid modeling approach for polymer melt index prediction

“…Rather than that, nonlinear machine learning modeling methods such as support vector machines (SVM), 12–13 least squares support vector machines (LSSVM), 15–17 relevance vector machines (RVM), 18 and Gaussian process regression (GPR) have been used 19–23 . SVM is a machine learning algorithm developed based on Vapnik–Chervonenkis (VC) theory and has been demonstrated good modeling performance in both classification and regression problems 24–26 .…”

“…Rather than that, nonlinear machine learning modeling methods such as support vector machines (SVM), [12][13] least squares support vector machines (LSSVM), [15][16][17] relevance vector machines (RVM), 18 and Gaussian process regression (GPR) have been used. [19][20][21][22][23] SVM is a machine learning algorithm developed based on Vapnik-Chervonenkis (VC) theory and has been demonstrated good modeling performance in both classification and regression problems. [24][25][26] GPR, also known as Kriging in geostatistics field, is a regression method based on an assumption that output variable is expressed as a regression function with a Gaussian prior distribution.…”

Soft Sensor Development Based on Quality-Relevant Slow Feature Analysis and Bayesian Regression with Application to Propylene Polymerization

“…Unfortunately, as the model is linear, it is inadequate when the soft sensing variables are nonlinear. To this end, nonlinear modeling methods are used in MI prediction, such as artificial neural networks (ANNs) [17], support vector machines (SVMs) [18], Gaussian process regression (GPR) [19,20], and relevance vector machine (RVM) [21] [ [22][23][24][25]. Recently, Liu et al proposed an adversarial transfer learning-(ATL-) based soft sensor [26] and a domain adaptation transfer learning soft sensor for product quality prediction [7].…”