Novel Framework for Simulated Moving Bed Reactor Optimization Based on Deep Neural Network Models and Metaheuristic Optimizers: An Approach with Optimality Guarantee.

Model-based optimization of simulated moving bed reactors (SMBRs) requires efficient solvers and significant computational power. Over the past years, surrogate models have been considered for such computationally demanding optimization problems. In this sense, artificial neural networks-ANNs-have found applications for modeling the simulated moving bed (SMB) unit but not yet been reported for the reactive SMB (SMBR). Despite ANNs' high accuracy, it is essential to assess its capacity to represent the optimization landscape well. However, a consistent method for optimality assessment using surrogate models is still an open issue in the literature. As such, two main contributions can be highlighted: the SMBR optimization based on deep recurrent neural networks (DRNNs) and the characterization of the feasible operation region. This is done by recycling the data points from a metaheuristic technique-optimality assessment. The results demonstrate that the DRNN-based optimization can address such complex optimization while meeting optimality.

Dynamics of a True Moving Bed separation process: Linear model identification and advanced process control.

The control of Simulated Moving Bed (SMB) units is challenging due to their complex dynamic behaviour and the difficulty of measuring their main properties. Furthermore, for the SMB units, the transfer function identification when the unit is operating at its optimal point is not easy to be done through the usual way. This work presents the development of a novel strategy to identify transfer functions of TMB/SMB and its application on classical linear model predictive controllers (MPC). However, for the process in study, due its unique dynamics, only the identification of the linear model is not enough to solve its control problem. Therefore, it is proposed a modification in the MPC prediction, that consists in a strategy based on a switching system where the most adequate transfer function is employed in the controller to overcome the problems related with the process dynamic behaviour. The results show that the used methodology enables the easy identification of transfer functions at the process optimal operating point and that the MPC can control the process in both the servo and regulator problem cases. It is also showed that the transfer function identified can be applied in the control of a SMB unit with four columns, under its optimal conditions.

A PSO-based optimal tuning strategy for constrained multivariable predictive controllers with model uncertainty.

This paper describes the development of a method to optimally tune constrained MPC algorithms with model uncertainty. The proposed method is formulated by using the worst-case control scenario, which is characterized by the Morari resiliency index and the condition number, and a given nonlinear multi-objective performance criterion. The resulting constrained mixed-integer nonlinear optimization problem is solved on the basis of a modified version of the particle swarm optimization technique, because of its effectiveness in dealing with this kind of problem. The performance of this PSO-based tuning method is evaluated through its application to the well-known Shell heavy oil fractionator process.