Importance of in-situ analysis in AI-driven control of hydrogen production

Gadi Briskman, Modcon Systems, Acco, Israel (gadib@modcon-systems.com)

There is a synergetic relationship between process control and process analysis. Our presentation will demonstrate the use case of hydrolyser operation as a perfect illustration of how online process analysis is as relevant as ever as we shift to the paradigm of AI-driven process control. Even more so, with the computational bottleneck of legacy control systems relieved, modern adaptive control systems can make the best use of real-time, precise readings of current state conditions.

One aspect of new data-driven modelling that seems to contradict the need for online analysis is the notion of digital twins. The digital twin is supposed to reproduce the process state given all necessary inputs over the specified time horizon. Digital twin technology has significantly benefited from recent advances in Machine Learning and AI algorithms. In fact, our approach to multivariate control integrates the digital twins as an inherent part of control model training and post-deployment updates.  

Yet there is a solid reason why the digital twin is not sufficient to ensure optimal control. This reason is the unmeasured disturbances.  In any process, there can be influencing factors that can not be effectively quantified or measured in real time. These factors cannot be among the process model's inputs. This means there will be a divergence between the digital twin's predictions and the actual values of the controlled variables: the rates and properties of the product streams.

The latter implicates two things. The first is that the need for direct measurements of the stream properties is not mitigated by advanced process modelling technology. The second is that the control system must leverage both the prognostic capability of the process model (digital twin) and the control model's ability to respond to changes in the controlled variables that deviate from predicted values.



Short Biography of Presenting Author

AI Business Development Manager at Modcon


Systems Ltd. He has several years of experience in computer vision, robotics, deep learning, and process optimisation. He holds an ME


degree in Systems Engineering and a BSc in Electrical Engineering.


Email: gadib@modcon-systems.com

Organized & Produced by:

www.isranalytica.org.il

POB 4043, Ness Ziona 70400, Israel
Tel.: +972-8-9313070, Fax: +972-8-9313071
Site: www.bioforum.org.il,
E-mail: hagit@bioforum.co.il