Final Thesis

ANN Artificial Neural Networks for Chemical Process Optimization

Key Info

Basic Information

Unit:
Process Systems Engineering
Type:
Masterthesis
Focus/Key Topic:
simulativ
Date:
2017

Contact

Description of the thesis:

In order to optimize complex chemical processes it is often necessary to develop detailed models of the process. The development of such models is often time-consuming and it can yield complex models, which are not useful for efficient optimization. Thus, the employment of machine learning surrogate-models such as artificial neural networks, Gaussian processes, and convex region surrogate models can be useful for optimization.

In the proposed work, the first step will be a literature review about common machine learning techniques. Then, promising techniques should be implemented and applied to a chemical process. One of the most important steps in the machine learning is the actual “learning” or parameter fitting. Thus, a focus of this work will deal with different methods of learning and will compare their performance. Finally, a chemical process should be optimized using the developed methods.

What we expect from you:

  • You should be open-minded and enjoy working in a team (also enjoy programming)
  • You should be interested in machine learning, neural networks, chemical processes, and optimization
  • You should have some basic knowledge about programming
  • It would be nice if you know already some Matlab, GAMS, gPROMS, and/or C++
  • You should be able to communicate in English

What we offer you:

  • A highly motivated team
  • Always a contact person for your work
  • Lots of challenging and interesting tasks
  • A possibility to gain in-depth knowledge about artificial neural networks, optimization, machine learning
  • If everything works well a possibility to continue your work as a HIWI at the institute

Contact:

If you have any further question or if you are interested in the work, please feel free to contact me. This work can possible be advised by a team three PhD students namely Adrian, Pascal, and Artur.

RWTH Aachen University

Aachener Verfahrenstechnik - Process Systems Engineering (AVT.SVT)

Forckenbeckstr. 51

Room A-203

52074 Aachen, Germany

Tel: +49 241 8097018

Email: artur.schweidtmann@avt.rwth-aachen.de

Web: http://www.avt.rwth-aachen.de