Model Calibration Method for Soft Sensors Using Adaptive Gaussian Process Regression

Wei Guo,Tianhong Pan,Shan Chen,Zhengming Li

doi:10.1109/access.2019.2954158

Abstract

The recursive Gaussian process regression (RGPR) is a popular calibrating method to make the developed soft sensor adapt to the new working condition. Most of existing RGPR models are on the assumption that hyperparameters in the covariance function are fixed during the model calibration. In order to improve the adaptive ability of the RGPR model, hyperparameters in covariance of Gaussian process regression (GPR) are adjusted in parallel by referencing the previous optimization. The matrix inversion formula is selectively used for updating the regression model. And a dynamic offset smoother is presented to further improve the reliability of the proposed method. Applications to a numerical simulation and the penicillin fermentation process evaluate the performance of the proposed method.

Highlights

As high requirements considering product quality, energy, safety in industrial processes, the numerical value of some certain necessary product indexes should be monitored in time
Large amounts of process relative data have been stored in the database by the distributed control system (DCS), which provides a reliable precondition to data-driven modelling [1], [2]
PCA and PLS project the multidimensional of the original variable space onto a low dimensional space through orthogonal principal components (PCs) and latent variables (LVs)

Summary

Introduction

As high requirements considering product quality, energy, safety in industrial processes, the numerical value of some certain necessary product indexes should be monitored in time. INDEX TERMS Gaussian process regression, hyperparameters-varying, model calibration, offset smoother, soft sensor. Support vector machine (SVM) [12], block-oriented nonlinear model [13], kernel PCA/PLS [14], Gaussian process regression (GPR) [15] have been widely used in the kernel functional method.

Results

Conclusion