Development of a data-driven model for the prediction of flood points of liquid-liquid extraction columns.
- Fluid Process Engineering
- Masterthesis / Bachelorthesis
- Focus/Key Topic:
Deep learning is a new field of machine learning, which has attracted much interest in recent years. It has been widely used in a variety of applications and has proven to be a powerful machine-learning tool for many complex issues.
In process engineering, various fields of application are currently tested and evaluated. In particular, for the design and determination of optimal operating points, artificial intelligence can play a major role.
At the AVT.FVT a data-driven model for the prediction of flood points of liquid-liquid extraction columns will be developed. For this, a database of approx. 2000 flood points is available. Based on this database, a data-driven model is to be developed that can predict the flood points independently of the column geometry and the multiphase system. The statistical quality of the model shall be compared with common flooding point correlations from the literature.
The work includes a literature research, the independent incorporation into Python, the optimization of the architecture of the network and the critical discussion of the results.
With this work you get:
- Profound knowledge in the field of artificial intelligence
- Insights into one of the most up-to-date and exciting topics in process engineering
- Optimization of operating conditions and the column geometry
- Deepening your ability to work scientifically
- Much freedom to contribute and implement your own ideas
- Experiences with simulative work
- Knowing Matlab, Python are helpful but not required
- Independent, structured and responsible way of working
- Motivation, interest and commitment
The thesis can be written in German or English. If you are interested you can contact me with a short curriculum vitae and an overview of your grades to make an appointment. E-Mail to: email@example.com