Master's Thesis: Robust Identification of Combined Electric Drive Models — Identifiability, Sensitivity & Excitation Analysis

At Bosch we shape the future with high-quality technologies and services that inspire enthusiasm and improve people's lives. Our promise to our employees stands firm: we grow together, enjoy our work and inspire each other. Welcome to Bosch. During your thesis you will delve into physics-based models of electric drives as well as our Python component library and simultaneously review current literature on robust system identification and experimental design (Design of Experiments). Furthermore, you will develop innovative robustness diagnostics for our existing identification pipeline – taking into account parameter identifiability, sensitivity and convergence behavior – and implement and quantify suitable countermeasures. Additionally, you will analyze the excitation content of existing data sets, evaluate relevant criteria and link these findings with the achievable identification quality. You will comprehensively evaluate the approach you developed using a practical benchmark use case. You will also document and present your research results clearly and understandably. Finally, you will extend the approach to address partially observable effects resulting from states that are not directly measurable, such as temperatures.

• During your thesis you will delve into physics-based models of electric drives as well as our Python component library and simultaneously review current literature on robust system identification and experimental design (Design of Experiments). Furthermore, you will develop innovative robustness diagnostics for our existing identification pipeline – taking into account parameter identifiability, sensitivity and convergence behavior – and implement and quantify suitable countermeasures. Additionally, you will analyze the excitation content of existing data sets, evaluate relevant criteria and link these findings with the achievable identification quality. You will comprehensively evaluate the approach you developed using a practical benchmark use case. You will also document and present your research results clearly and understandably. Finally, you will extend the approach to address partially observable effects resulting from states that are not directly measurable, such as temperatures.

• Education: Master's degree in Cybernetics, Engineering, Mathematics, Computer Science or equivalent
• Experience and Knowledge: Basic knowledge of dynamic systems and differential equations; very good knowledge of Python programming; understanding of physics of electric machines; optional: experience with MATLAB/Simulink for interaction with existing simulation models; optional: knowledge in Machine Learning and frameworks for automatic differentiation (JAX or PyTorch)
• Personality and Work Style: You work independently and systematically on complex issues and have strong analytical skills, enabling you to precisely capture and solve problems
• Work Mode: Office presence required
• Languages: Very good English skills