Data‐driven model predictive quality control of batch processes

Abstract

The problem of driving a batch process to a specified product quality using data-driven model predictive control (MPC) is described. To address the problem of unavailability of online quality measurements, an inferential quality model, which relates the process conditions over the entire batch duration to the final quality, is required. The accuracy of this type of quality model, however, is sensitive to the prediction of the future batch behavior until batch termination. In this work, we handle this "missing data" problem by integrating a previously developed data-driven modeling methodology, which combines multiple local linear models with an appropriate weighting function to describe nonlinearities, with the inferential model in a MPC framework. The key feature of this approach is that the causality and nonlinear relationships between the future inputs and outputs are accounted for in predicting the final quality and computing the manipulated input trajectory. The efficacy of the proposed predictive control design is illustrated via closed-loop simulations of a nylon-6,6 batch polymerization process with limited measurements.

Self-tuning final product quality control of batch processes using kernel latent variable model

“…It should be noted that, the multi-PLS models based approach proposed by Wang and Srinivasan (2009) built a finite set of quality models at each decision points, and data only up to the decision point is used. However, these quality models developed at early time points may be highly inaccurate because they will not capture the effects of large periods of the batch duration toward the batch quality (Aumi et al, 2013). Hence, the control performances of multi-PLS model are much poorer than that of PLS model with adding soft constraint on score magnitude and missing data imputation method.…”

“…model building and future trajectories estimation, and the former is not essentially dependent on the latter and vice versa. Hence it is not necessary to use the same model at different stage (Aumi et al, 2013;Zhang and Qin, 2007;Cho, 2007). And the prediction accuracy can be increased by improving either of the two stage or both.…”

“…by using multiple models (Zhang and Hu, 2011) or a different unfolding scheme (Zhao et al, 2009). However when using the model for control, the prediction of the future trajectories for a given manipulated variables is a necessity (Aumi et al, 2013). Fortunately, the final quality prediction can be readily divided into two separate stages, i.e.…”

Multivariate batch to batch optimisation of fermentation processes incorporating validity constraints

“…Further studies in to the use of multivariate statistical models to regulate batch processes have been conducted by several research groups, including [9][10][11][12][13]. Whilst this work has been demonstrated to be successful, multivariate techniques are not without their limitations.…”

Personalized Forecasting System for Complex Product Quality Based on Predictive Control Model of Multiple Linear Regression Analysis

“…On the basis of a lot of experimental data, experimental mathematical models of predictive control model of several sorts of fault characteristics are built up with multiple linear regression analysis [15]. Using predictive control model of multiple linear regression analysis, product quality has been forecasted synthetically [16].…”