Optimal Lap in Racing: Custom PDE Solver and Problem Formulation for Raceline Optimization for Minimum Lap Times

Institut

Lehrstuhl für Fahrzeugtechnik

Typ

Semesterarbeit / Masterarbeit /

Inhalt

theoretisch /  

Beschreibung

Imagine developing cutting-edge optimization algorithms that directly impact the performance of real, high-speed,
autonomous racecars. This thesis offers the unique opportunity to design and implement a custom partial differ-
ential equation (PDE) solver for 3D minimum laptime optimization, a critical component in maximizing the speed
and efficiency of autonomous racecars in three-dimensional space. As part of this project, you will be a key contrib-
utor to the TUM Autonomous Motorsport (TUM AM) team—one of the most successful autonomous racing teams
in the world. TUM AM recently won the inaugural Abu Dhabi Autonomous Racing League (A2RL) event
and has consistently been a top contender in prestigious competitions like the Indy Autonomous Challenge (IAC).
Your work will not only advance state-of-the-art offline trajectory optimization but will also be directly integrated into
the software stack of a real, open-wheeled autonomous formula racecar, shaping its competitive edge on the track.
The ability to compute an optimized racing line is fundamental to autonomous racing success. This thesis focuses
on the development of a custom PDE solver tailored for minimum lap-time optimization in complex, high-speed
racing scenarios. The task is currently solved with a Nonlinear Program generated from direct collocation of an
Optimal Control Problem with RK4, which is then solved with a generic solver like IPOPT. In this thesis the goal
is to improve convergence and allow for more complex vehicle representations by either creating a custom solver
tailored to this application or changing the problem formulation for more efficient solving. The first step will be to
investigate existing state-of-the-art methods for minimum lap-time optimization, analyzing different PDE-based ap-
proaches to understand their advantages and limitations. Based on this analysis, a novel PDE solver will be de-
signed to accurately capture the vehicle limits in high-speed racing. A some added motivation to start strong from
day one, the faster this basic goal is satisfied the more time there will be for you to integrate additional sources of
complexity: maybe the effects of curbs or explicit aerodynamics, you decide!

Voraussetzungen

• Enthusiasm for racing
• Good programming skills in Python or C++ (preferably both, high level framework and visualizer in python + fast optimizer in a compiled language)
• Ability to collaborate in a team and engage in interdisciplinary research

Tags

FTM Studienarbeit, FTM Automatisiertes Fahren, FTM Buettner, FTM Informatik

Möglicher Beginn

sofort

Kontakt

Sascha Büttner, M.Sc.
sascha.buettnertum.de

Ausschreibung