Numerical Thermo-Fluids - From Differential Equations to Deep Learning

What it is about

You will learn numerical algorithms and good programming principles to solve thermo-fluid dynamic and thermal engineering problems. In addition to classical numerical methods, you will receive an introduction to Machine Learning and Deep Learning. In the end you can implement basic algorithms in MATLAB or use existing MATLAB functions and create and train simple neural networks with the Deep Learning Toolbox.

What you will be able to do

• Implement numerical algorithms in MATLAB
• Apply existing MATLAB functions and tools
• Understand fundamental concepts of Deep Learning
• Create and train neural networks with MATLAB Deep Learning Toolbox
• Understand solution approaches for selected thermo-fluid dynamic tasks and transfer them to analogous engineering problems

What the module consists of

• VorlesungConveying the theoretical foundations of numerical computation, ODEs, PDEs, Runge-Kutta, Machine Learning and Neural Networks
• Praktische Übungen / ProgrammieraufgabenImplementation of the lecture content by implementing the algorithms in MATLAB
• AbschlussprojektApplication of what has been learned to a larger topic with a report and presentation

Teaching method

• Vorlesungsfolien und SkripteConveying the theoretical foundations
• Bereitstellung von Materialien über MoodleAccess to teaching materials and exercise documents
• Programmierung in MATLAB nach jeder VorlesungImmediate practical application and consolidation of the contents