Upcoming Presentations

Master Thesis

 

Abstract

As part of this master’s thesis, a precise cell holder for large-format pouch cells was developed, providing both mechanically stable fixation and reliable electrical contacting for cell testing. The aim was to design a holder structure that ensures defined and uniform compression of the cell while simultaneously enabling a low-resistance connection between the device under test and the measurement system. Particular attention was paid to the galvanic separation of load and measurement lines in order to minimize parasitic effects and significantly improve measurement accuracy.

The entire holder was electrically insulated to ensure safe operation in electrically conductive environments, especially in climate chambers. The compression force applied to the cell is generated by an external press and continuously monitored using a custom-designed NTC board. Additionally, the system allows the integration of supplementary sensors for measuring internal pressure, mechanical deformation via LVDT sensors, and temperature distribution. This enables comprehensive monitoring of cell stress and precise analysis of the electromechanical behavior of the pouch cells.

Master Thesis

 

Abstract

Accurate modeling of lithium-ion battery systems is a key prerequisite for developing high-performance battery management systems (BMS) in electric vehicles and stationary energy storage. This master’s thesis develops, parametrizes, and validates a real-time capable 3RC Thevenin model for a 12s2p lithium-ion battery module. The model is based on an electrical equivalent circuit with three RC branches representing electrochemical polarization processes on different time scales.

Parameter identification is performed using the least squares method based on stepwise discharge tests at four temperatures (0°C, 10°C, 25°C, 40°C) and a discharge rate of ⅓ C. The open-circuit voltage (OCV) and RC parameters are extracted from relaxation phases and stored in temperature- and SOC-dependent lookup tables. Validation demonstrates high model accuracy (RMSE = 30.7 mV, MRE = 0.04%) and confirms generalization capability through additional tests with ½ C discharge and a WLTP load profile (MRE = 0.11%, max. error = 0.49%).

By scaling the model to cell level, a 12-cell configuration is derived that accurately reproduces individual cell voltages (RMSE = 26.8 mV, MRE = 0.03%) and enables detection of faulty cells. The developed model thus provides an excellent balance between accuracy and computational efficiency and is well suited for model-based SOC estimation and the analysis of inhomogeneous battery systems.