Location: JHE 326H
The aim of this talk is to present the recent tools developed in my research group for multi-scale process systems under uncertainty and integration of design and control. Thin film deposition is an important process that can be modeled using macro-scale continuum equations embedded with micro-scale lattice-based kinetic Monte Carlo simulations. Strong dependence of the electrical and mechanical properties of thin films on their microstructure has motivated research on the control of thin film growth. From the modeling point of view, the evolution of the thin film encompasses the microscopic processes that are subject to model parameter uncertainty. To achieve the desired thin film’s characteristics, it is crucial to develop control strategies that are robust to these uncertainties. Although there are devices that can measure system’s micro-scale properties, most of these measurements are not available as frequent as required for the development of an effective feedback control strategy. Moreover, the lack of a closed formulation between the controlled outputs and the model parameters increases the problem’s complexity. Motivated by these observations, this talk will address the method recently developed in my group to analyze model uncertainty propagation employing tools such as Taylor Series Expansion. The effectiveness of the proposed tool has been evaluated and used to perform open-loop control on the thin film deposition process. The results from this analysis reveal that model uncertainty significantly affects the fine-scale properties of multi-scale systems and hence it needs to be taken into account for the controllability of these processes.
Integration of design and control aims to address the optimal design while simultaneously performing a controllability analysis on the chemical system. In this talk, recent approaches developed by my group to address the simultaneous design and control will be presented. The first approach presents an iterative scheme that includes robust dynamic feasibility and stability analyses, which are aimed to ensure the process dynamic operability and asymptotic stability under uncertainty. These analyses are formulated as convex mathematical problems, which makes this method computationally attractive. The second approach aims to address the integration of design and control for disturbances that follow a probabilistic-based description while using advanced model-based control schemes such as Model Predictive Control (MPC). The key idea is to determine the dynamic variability of the system that will be accounted for in the process design using a stochastic-based worst-case variability index. Case studies involving a ternary distillation unit and an actual wastewater treatment industrial plant are presented and used to test the effectiveness of the proposed methodologies.
Return to list of all seminars