Master Thesis: Combining MPC and AI on Motion Simulators

What to expect

In the department of flight control and simulation one of the main activities is the development of full motion simulators and algorithms to control them. The algorithms consist of two parts: the first part is called Motion Cueing Algorithm that transfers the motion from the simulation (e.g. flight simulation or driving simulation) to the workspace of the motion platform. The second leads to the actual movement of the platform by providing joint angles or a Cartesian pose that is processed by the controller of the platform. The goal of the thesis is to combine a Model Predictive Control (MPC) approach with AI techniques to improve the movement significantly.

Your tasks

• Familiarize with the current setup of the motion platform (mechanical, system dynamics, control)
• Familiarize with the current setup of a MPC controller including a human perception model of the vestibular system
• Research on human perception and implementing a metric to measure the quality of the motion simulation
• Applying Reinforcement Learning to improve the MPC approach by changing parameters according to the current simulation (online)

Your profile

• Bachelors degree in mathematics, informatics or a comparable study programme
• Very good programming skills in Python
• Knowledge in AI
• Being able to work responsibly and independently