Real-Time Optimization for Predictive Thermal Management of Electric Vehicles

Background
Traton is developing the next-generation of advanced thermal management systems for electric vehicles. In electric vehicles, the thermal management of batteries is an essential component of the software functionalities to run the vehicle safely and efficiently. The planning of the upcoming driving route or a charging session provides us with valuable information about the future power demands on the battery, which enables us to anticipate the thermal load on the battery to optimize the thermal management of the vehicle. Predictive thermal management in electric vehicles relies on solving optimization problems that translate future scenarios into actionable control strategies. This Master Thesis will focus on the real-time optimization solution that determines the actuator commands (e.g., cooling systems, battery heaters), making it essential to balance performance, feasibility, and robustness. The work will be carried out in parallel with another thesis focused on defining and quantifying control objectives. Together, these projects will form a comprehensive foundation for the development of predictive thermal control strategies in electric vehicles.

Objectives

• Solve the optimization problem, based on predefined control goals and constraints; including system dynamics, actuator limits, and operational boundaries

• Explore methods suitable for real-time execution (e.g., simplified MPC, convex approximations, surrogate models) and evaluate trade-offs between accuracy and computational cost

• Ensure optimization results can be translated into actionable commands for thermal actuators, considering both direct control (e.g., fan speed) and indirect strategies (e.g., setpoint adjustments)

• Simulate and validate the solution under realistic driving and charging scenarios

• Assess robustness to disturbances, uncertainties, and missing data

• Provide recommendations for implementation, highlighting limitations and areas for future improvement

The thesis will focus on:

• Real-time optimization solution for predictive thermal management

• Translation of the optimization output into actionable commands for thermal actuators

• Evaluation of the feasibility, robustness, and computational efficiency of the solution under realistic driving and charging scenarios

It will not involve hardware implementation or low-level controller design.

Job description
The work will include:

• Literature review

• Propose and develop computationally efficient optimization algorithms

• Translate optimization outputs into actuator-level commands

• Simulate and validate the solution

• Analysis of the results, evaluating feasibility, robustness, and scalability

• Summarize findings and provide recommendations for real-time implementation

You will work in a dedicated team of driven experts from many aspects of batteries, charging and thermal management of the vehicle. The thesis will provide theoretical support and validation for future integration.

Education/program/focus
Master’s student in Electrical Engineering, Energy Systems, Vehicle Engineering or equivalent with a background in control theory, automotive systems, or applied mathematics with strong skills in optimization, control systems, and simulation. Experience with simulation tools (e.g., MATLAB/Simulink, Python) will be appreciated. You should have an interest in electric vehicle, battery technology and control.

Number of students: 1
Location: Traton R&D, Södertälje, Sweden
Start date: January 2026
Duration: ~ 20 weeks (Full-time)
KTH Supervisor: Jörgen Wallin, jorgen.wallin@energy.kth.se
Scania Supervisor: Göran Lissel, goran.lissel@scania.com