Challenges in Forecasting Membrane Fouling in Filtration Processes using Univariate Data-Driven Models

Membrane-filtration processes play a crucial role in modern pharmacy and medicine but also in food industry and in biotechnology. During filtration, the filtered particles increasingly block the filter membrane leading to a reduced throughput up to a total blockage over time. Forecasting this membrane fouling would allow appropriate countermeasures to be taken in time to slow down the blockage and thus increase the overall yield. This paper introduces the challenges of yield maximization in membranefiltration processes despite membrane fouling and possible advantages in forecasting membrane fouling. Further, the paper discusses the difficulties in the membrane-filtration process data and its pre-processing which need to be addressed to realize forecasting membrane fouling using data-driven machine learning and data analytics in process control     «

Membrane-filtration processes play a crucial role in modern pharmacy and medicine but also in food industry and in biotechnology. During filtration, the filtered particles increasingly block the filter membrane leading to a reduced throughput up to a total blockage over time. Forecasting this membrane fouling would allow appropriate countermeasures to be taken in time to slow down the blockage and thus increase the overall yield. This paper introduces the challenges of yield maximization in memb...     »