A Dynamic Mode Decomposition Predictive Control scheme for an Osmotic Membrane Distillation Process

In this paper, a Data-Driven predictive control strategy is proposed for the regulation of the transmembrane flux in an Osmotic Membrane Distillation (OMD) lab-scale unit operating with a sucrose solution as feed and MgCl2 as hypersaline draw solution. The control objective is to accurately track desired flux setpoints under the intrinsic nonlinearities of the underlying mass and heat transport phenomena. To this end, physicochemical laws are combined with experimental data to develop a digital twin of the process. Such digital twin allows the collection of simulated informative data to compute a linear description of the plant system in a data-driven fashion. In particular, the proposed approach is based on the Dynamic Mode Decomposition with Control (DMDc) algorithm. The identified model is then embedded within a predictive control scheme, allowing the constrained regulation of the unit. Numerical simulations demonstrate the effectiveness of the proposed method in driving the system toward multiple operating points while ensuring accurate flux regulation.