Data-Based Modeling and Control of Nylon-6, 6 Batch Polymerization

Abstract

This work addresses the problem of modeling the complex nonlinear behavior of the nylon-6, 6 batch polymerization process and then subsequently tracking trajectories of important process variables, namely the reaction medium temperature and reactor pressure, using model predictive control. To this end, a data-based multi-model approach is proposed in which multiple local linear models are identified from previous batch data using latent variable regression and then combined using an appropriate (continuous) weighting function that arises from fuzzy c-means clustering. The proposed approach unifies the concepts of auto-regressive exogenous (ARX) modeling, latent variable regression techniques, fuzzy c-means clustering, and multiple local linear models in an integrated framework capable of capturing the nonlinearities and multivariate nature of batch data. The resulting data-based model is then used to formulate a trajectory tracking predictive controller. Through simulation studies, the modeling approach is shown to capture the major nonlinearities in the nylon-6, 6 polymerization process and closed-loop simulation results demonstrate the efficacy of the proposed predictive controller and illustrate its advantages over existing trajectory tracking approaches such as conventional proportional-integral control and latent variable model predictive control.