We are the Chair of Vehicle Mechatronics at Technische Universität Dresden, specializing in battery research. Our work addresses key challenges in battery engineering, such as modelling the electrical, thermal, and ageing behaviour based on laboratory and fleet data, using various modelling approaches (empirical, data-driven, and physical).

We offer topics in this field across multiple disciplines, including electrical, mechanical, and software engineering.

Physics-based battery models such as the Doyle–Fuller–Newman model (DFN) realistically represent electrochemical and thermal processes inside a cell, enabling explainable predictions of voltage, temperature, concentration profiles, and aging mechanisms. They are therefore a key tool in battery research and development.

The challenge: Such models contain a large number of physical parameters, many of which are temperature- and SOC-dependent. Directly optimizing all parameters at once is practically infeasible due to the high number of degrees of freedom and the resulting over-determination. Traditional methods rely on time-consuming chemical analyses and individual measurements— often with high effort and limited transferability to new cell chemistries.

This work pursues a data-driven approach that combines existing test bench measurements (EIS, pulse tests, OCV at various temperatures and SOCs, optionally with a reference electrode) with methods from sensitivity analysis and machine learning. The goal is to derive effective parameter blocks from the measurement data, enabling robust and efficient model adaptation in PyBaMM (Python Battery Mathematical Modelling)—without the need for elaborate chemical detail measurements.

Focus of the Thesis

1. Data Preparation

• Consolidation, verification, and structuring of existing EIS, pulse, and OCV data
• Segmentation according to frequency ranges and load profile

1. Sensitivity Analysis & Parameter Reduction

• Identification of the most influential model parameters
• Grouping of multiple physical parameters into optimizable effective parameter
  blocks

1. Parameter Optimization

• Development and application of suitable optimization strategies (e.g., gradientfree
  + gradient-based methods)
• Ensuring physical plausibility and consistency conditions
• Research into concepts/measurement methods for further differentiation of
  parameter blocks where applicable

1. Implementation in PyBaMM

• Creation of a custom parameter set
• Adaptation and validation of the model within PyBaMM

More Information:
https://drive.google.com/file/d/1Nj0XUKmzoV2b5jPiqpBMCYQ7n92hWZLh/view?usp=sharing

• Interest in scientific research
• Interest in modelling across various domains
• Willingness to learn new topics (e.g., battery technology)

Please send your application, including a CV and your transcript of records, to the mentioned contact. We’ll be in touch.