PhD position – Hybrid Physics-Neural Network Soft Sensors for Dynamic Process Operation (DFG...

Work group:

IEK-10 - Modellierung von Energiesystemen

Area of research:

PHD Thesis

Job description:

In the Industrial Energy Systems group at the Institute of Energy and Climate Research – Energy Systems Engineering (IEK-10) we develop models, methods, and algorithms to optimally design and operate future industrial energy systems. The shift from a consumer-driven to a producer-driven energy system is both an opportunity and a challenge for energy-intensive industries. By means of intelligent system integration and automation (digitalization), direct or indirect electrification (Power-to-X), and flexible operation, industrial processes can reduce their environmental impact and energy cost and thus make an important contribution to a successful Energy Transition.

Your Job:

The advertised PhD position relates to a new project that is part of the DFG Priority Programme “Machine Learning in Chemical Engineering” (SPP 2331, chemengml.org ) funded by Deutsche Forschungsgemeinschaft (DFG, German Research Foundation). For this project, we at the IEK-10 have teamed up with the Chair for Fluid Process Engineering (AVT.FVT) at RWTH Aachen University. The goal of the project is to explore the potential of hybrid physics-neural network models, particularly so-called physics-informed neural networks (PINNs), as a novel approach for inferring poorly measurable states from liquid-liquid separation processes for which mechanistic descriptions are only partially available. Specifically, we will study the inference of the dispersion layer height in a horizontal, continuously operated gravity settler, a common low-energy separation unit for liquid-liquid dispersions in chemical, biotechnological, and recycling processes. To this end, our project partners at AVT.FVT will generate experimental data and provide mechanistic knowledge as a basis for development, training, and validation of the hybrid physics-neural network models that form the core of this PhD project. Your tasks in detail:

• Develop a concept for PINN-based hybrid physics-neural network soft sensors
• Implement, train, and validate a hybrid model for operation of a gravity settler under varying operating conditions
• Compare the hybrid model to a fully data-driven benchmark, e.g., a recurrent neural network (RNN)
• Develop, implement, and validate a real-time capable approach for state estimation
• Assess the validity range of the hybrid model

Your Profile:

• Excellent Master’s degree in computational engineering, process systems engineering, simulation science, or a relevant discipline
• Very good knowledge in both machine learning and process systems engineering
• Expert knowledge of at least one programming language (preferably Python)
• Excellent skills in spoken and written English
• Excellent organizational skills and the ability to work independently
• Excellent cooperation and communication skills and ability to work as part of a team

Our Offer:

We offer ideal conditions for you to complete your doctoral degree:

• Possibility of pursuing a doctoral degree (Dr.-Ing.) at RWTH Aachen University under the supervision of Prof. Alexander Mitsos
• A highly motivated research group in one of the biggest research centers in Europe working on optimization and machine learning for design and operation of energy processes/systems
• Close cooperation with our partners at the Chair for Fluid Process Engineering (AVT.FVT) and the Institute of Process Systems Engineering (AVT.SVT) at RWTH Aachen University
• An excellent scientific and technical infrastructure: both necessary conditions for a successful PhD thesis at RWTH Aachen within three and a half years
• Participation in project meetings and conferences
• Strong support and mentoring for setting up a future career in science and / or the industry

A structured doctoral degree program with a comprehensive further training and networking package (see our doctoral researchers’ platform JuDocs www.fz-juelich.de/judocs )

• Targeted services for international employees, e.g., through our International Advisory Service

The employment of doctoral researchers at Jülich is governed by a doctoral contract, which usually has a term of three years. Pay in line with 75% of pay group 13 of the Collective Agreement for the Public Service (TVöD-Bund) and additionally 60 % of a monthly salary as special payment („Christmas bonus“). Further information on doctoral degrees at Forschungszentrum Jülich including our other locations is available at: www.fz-juelich.de/gp/Careers_DocsForschungszentrum Jülich promotes equal opportunities and diversity in its employment relations.We also welcome applications from disabled persons.

This research center is part of the Helmholtz Association of German Research Centers. With more than 42,000 employees and an annual budget of over € 5 billion, the Helmholtz Association is Germany's largest scientific organisation.