Skip to main content

Smart tomographic sensors for advanced industrial process control

Periodic Reporting for period 1 - TOMOCON (Smart tomographic sensors for advanced industrial process control)

Reporting period: 2017-09-01 to 2019-08-31

With the most recent progress in fast data processing on smart scalable parallel hardware architectures imaging techniques have reached the capability of being employed as sensors in advanced real-time control systems. A decisive gap in the process industry is however the lack of qualified distributed process parameter measurements for optimal control. Process tomography is a ground-breaking technology to bridge this gap as it can give insights into opaque process components and materials.

The European Training Network “Smart tomographic sensors for advanced industrial process control (TOMOCON)” joins 12 international academic institutions, 15 industry partners and 15 early stage researchers who work together in the emerging field of industrial process control using smart tomographic sensors. The project shall generate new scientific and technical knowledge in this emerging field; develop and demonstrate new technological solutions of advanced industrial control by tomographic sensors and align tomography-based process control with the concepts of big data analysis, advanced human-machine interfaces, knowledge-based control as well as sensor integration into a networked production according to the future needs of the industrial associated partners.

Together with world-leading industrial sensor and control solution providers as well as process engineering and production companies, TOMOCON shall demonstrate the functionality of tomography-assisted process control in four industry-relevant demonstration cases that are inline fluid separation, microwave drying, continuous metal casting and batch crystallization. These demonstrations will serve as benchmarks to demonstrate improvements in energy and resource efficiency as well as product quality in near industrial environments.

TOMOCON will further educate its 15 early stage researchers in a new way that combines unique scientific and soft skill training based on intersectoral, international and interdisciplinary experiences to enable technical breakthroughs and innovations in Europe.
A variety of tomographic sensors has been developed to fit the different requirements and control objectives for the four demonstration processes: Electrical resistance tomography (ERT) and wire-mesh sensors (WMS) for inline fluid separation, electrical capacitance tomography (ECT) and microwave tomography (MWT) for microwave drying, contactless inductive flow tomography (CIFT) and magnetic induction tomography (MIT) for continuous casting and ultrasound tomography (UST) and ERT for batch crystallization.

For each process the appropriate numerical methods have been selected resulting in the choice of different numerical tools: The in-house JADIM code for inline fluid separation, COMSOL Multiphysics for microwave drying, OpenFOAM for continuous casting and ANSYS Fluent for batch crystallization.

Furthermore, preliminary controller structures have been achieved for all four demonstrations. All of these structures have been implemented as virtual (simulated) controllers using the standard MATLAB/SIMULINK environment. For batch crystallization and for the inline flow swirl separator, simplified computational fluid dynamics models were used and certain physical phenomena were simplified to achieve a lumped parameters single-input single-output model. Similarly for microwave drying, the partial differential equations describing the process have been space discretized using the finite elements method. These methods resulted in a lumped parameters model in a state space form. In the case of continuous casting, raw data from the tomographic sensors was used and processed to obtain suitable numerical characteristics related to the product quality. A dynamic relationship between the control actions and these characteristics was obtained using system identification methods. These lumped parameters models were used as the starting points for the design of the controllers using a variety of approaches: Proportional Integral Derivative controller (PID), Linear Quadratic Regulator (LQR) and Model Predictive Control (MPC).

For inline fluid separation, a virtual demonstrator was built using a simple physically-based model and tomographic reconstruction algorithm, based on synthetic flow fields, combined with WMS and ERT. For microwave drying, a virtual demonstration tool was also developed and tested. It comprises a PDE based process model, virtual ECT and MWT sensors, forward and inverse solvers, model-based control concepts and some visualization tools. In the case of the virtual demonstrator for continuous casting, a virtual CIFT as well as a virtual MIT/ECT sensor were implemented. In summary, the demonstrator is running and a variety of single-phase UDV measurements exist. For batch crystallization, a virtual process control model has been developed by the use of virtual data from 1-D tomographic sensors.
The TOMOCON research and training programme intents to greatly enhance the professional expertise and the scientific and transferable skills of the ESRs. At the end of the project, each ESR will acquire a full understanding of the comprehensive approach of the TOMOCON network, ranging from improved sensor design and materials, advanced data analysis from flow field measurements, complex modelling and simulation as controller input to high-tech experimental tools and four distinguished demonstrations based on the clearly identified needs of important branches of the process industry. For this, TOMOCON has created a broad network that involves several highly ranked companies along the value chain, from European SMEs to world leading engineering providers and global producers as well as four renowned international advisors from the UK, USA, China and Japan.

To improve the career perspectives and expertise of the ESRs, TOMOCON is offering visits and training in the network partners’ top level research facilities. By the end of August 2019, all ESRs conducted in total 30 secondments at partner institutions and companies and gained a deeper understanding of industrial processes and research activities during company visits to the industrial partners Siemens, Linde, Netrix, Tata Steel and Shell.

Furthermore, transferable soft skill courses will broaden the career prospects of the ESRs by enhancing their competences and awareness towards production, leadership, business planning, IPR creation and entrepreneurship.

The project will also create a remarkable attention in the scientific community by achieving a significant number of scientific publications in highly-ranked scientific journals and presentations at renowned international conferences and workshops.

As research on advanced process control is emerging worldwide, particularly in the US and Asia, we are at a point when technological breakthroughs in novel tomographic sensors and control strategies are needed to secure Europe's leading position in this field. Considering the partners involved, TOMOCON coordinates key European contributions and forms a world leading network in this area. Therefore TOMOCON has recruited 15 top level early stage researchers from around the world. The large involvement of the private sector and leading research institutions will furthermore significantly enhance the research and innovation capacity of all participating organizations and form a critical mass needed within Europe to create innovation in the form of new products and services.
TOMOCON Workshop Delft, 27-28 June 2019
TOMOCON Summer School Lublin, 26-28 September 2018
WCIPT-9, 2-6 September 2018
TOMOCON Kick-Off Meeting Dresden, 17-20 April 2018
Company Visit Linde, 20 April 2018
Company Visit Tata Steel, 22-23 November 2018
TOMOCON Workshop Lodz, 24-25 September 2018
TOMOCON Control Workshop Liberec, 19-24 September 2019
TOMOCON Summer School Delft & Company Visit Shell, 1-3 July 2019
Company Visit Netrix, 26 September 2018
Company Visit Siemens, 19 April 2018