Postdoctoral researcher

Last application date
Feb 01, 2020 14:52
Department
TW08 - Department of Electromechanical, Systems and Metal Engineering
Employment category
Postdoctoral staff member group 3
Contract
Limited duration
Degree
PhD in electromechanical or control engineering
Occupancy rate
100%
Vacancy Type
Research staff

Job description

Postdoctoral researcher on deep learning predictive control for industrial processes.

Postdoctoral fellowship vacancy for Moonshot project ‘Control algorithms for flexibility in power-to-X and industrial processes’. We are particularly looking for a team member with a strong background in system identification and optimal control for dynamical systems. Starting date is as soon as possible, preferably 1st March 2020. For this position, we offer an internationally competitive salary that depends on seniority and corresponds to the salary scales for Postdoctoral Research Fellows as established by the Flemish government.

Summary of project:
The general goal of the overall project is to unlock and control the flexibility of the energy-intensive processes and of power-to-X to optimally reduce the CO2-emissions in a way that is best suited for the overall industrial and energy eco-system. In this project, innovative methods need to be researched and initiated to design and elaborate deep-learning-based techniques to optimally control the flexibility over multiple energy vectors, according to different objective functions.

The fundamental research challenge within the project lies in the design of new techniques that combine modern data-driven methods with optimal control approaches to address the challenges faced by the energy-intensive industry. The specific research challenge lies in tackling the highly uncertain industrial system that is difficult to model and if a model is available it might be non-differentiable further aggravating finding optimal control policies. Hybrid AI approaches will be furthermore researched by using reinforcement learning for policy refinement.

The research group on dynamics for electromechanical systems in which you will be embedded has extensive expertise in the research on hybrid AI methods for system identification/intelligence and optimal control that unlock functionalities to take AI to the real world.

We offer to you:

- The research group is situated in Ghent, a lively city at the heart of Europe (https://visit.gent.be/en/HOME).

- The candidate will have access to state-of-the-art tools and facilities, a network of Flemish companies active in the manufacturing industry, and the possibility to collaborate with other research groups.

- The candidate will be directly embedded in an international research group of +20 people that work together as a team and will have the possibility to collaborate with many other people active at UGent both associated with Flanders Make and IMEC.

- Starting date: as soon as possible, but flexible.

- You will be offered a contract of 1.5 years that can be extended.

Profile of the candidate

You have a master of science degree in engineering, preferably in electromechanical or control engineering and hold a PhD in electromechanical/control engineering. You are motivated, creative and have a passion for academic research. You have excellent publication record, with publications in international peer-reviewed academic journals. You have excellent spoken and written academic English.

You have extensive experience with the modelling and simulation of dynamical systems, preferably of industrial processes. You have extensive experience with machine learning (supervised learning, reinforcement learning) as well as with system identification and nonlinear optimal control techniques for industrial processes.

The candidate will be expected to:

- Perform research on deep learning optimal control for industrial processes. You will set out a technological basis for combining optimal control theory with reinforcement learning techniques.

- You will coordinate activities related to the control of industrial processes and interact with PhD student.

- Participate to team meetings and external meetings with other research partners.

- You will look at opportunities to transfer the knowledge gained towards mechatronic and robotic systems.

- You will cooperate with researchers active within the research group and outside.

- You will be embedded in an internationally competitive research group that has a strong focus on bringing artificial intelligence to the real world.

- You will map your developed approaches with the industrial needs and initiate collaborations with industry, other research groups and research institutions, both within Flanders as on the European level.

How to apply

Send your CV and motivation letter to Guillaume.Crevecoeur@UGent.be preferably before the 1st of February 2020 with subject ‘post doc dControl’. Please include contact details of two referees. If you pass the pre-selection, you will receive further instructions on the selection process and will be invited for a job interview, possibly via Skype.