Oscar Palma Flores

I received my BASc in Chemical Engineering from the Benemerita Universidad Autonoma de Puebla (Mexico) in 2010, a MSc degree in Chemical Engineering from Universidad Iberoamericana (Mexico), and currently pursuing a PhD degree at the University of Waterloo. My research is focused on the integration of design and model-based control under uncertainty. I implement modern control strategies such as nonlinear model predictive control (NMPC). Multiple challenges arise for the solution of the corresponding optimization formulation for design and NMPC-based control:  a) a multilevel formulation (embedded optimization problems), b) introduction of uncertainties and disturbances, c) implementation of structural decision using integer variables. In particular, I have implemented two different strategies for the solution of the multilevel formulation: (i) a decomposition strategy based on the determination of the back-off that allows to determine a dynamically feasible process design [1], and (ii) a direct strategy based on the transformation of the multilevel formulation into a single-level problem using a classical KKT transformation [2]. On the other hand, the introduction of structural decisions increases the complexity of the problem. To circumvent this challenge, I have implemented a discrete-steepest descent algorithm to determine the optimal structure, process design, and control.

[1] Palma-Flores, O., & Ricardez-Sandoval, L. A. (2022). Simultaneous design and nonlinear model predictive control under uncertainty: A back-off approach. Journal of Process Control, 110, 45-58.

[2] Palma-Flores, O., & Ricardez-Sandoval, L. A. (2022). Integration of design and NMPC-based control for chemical processes under uncertainty: An MPCC-based framework. Computers & Chemical Engineering, 162, 107815.