1 October 2020, 11.15-12.00
Daniel Luder, former Master’s student at BFH (now working at evolaris aviation GmbH and ALGOnovum GmbH)
Accurate State Estimation of Lithium-Ion Batteries using Physics-Based Models
The increasing energy and power density of lithium-ion batteries introduced electric vehicles (EV) to the mass market. Besides the lithium-ion cells, the battery management system (BMS) with its battery models and algorithms has become a core component of the entire battery system. This webinar focuses on a new method by using pseudo-2D physics-based battery models based on theories of porous electrodes and concentrated solutions with the advantage of expressing electrochemical meaningful states within the cell that are essential for fast charging and degradation minimization approaches. It is shown how the most sensitive model parameters can be identified for a commercial lithium-ion cell by microstructure analysis and multi-objective model optimization. The model is validated with drive cycles as they occur in EVs revealing a RMSE smaller than 18mV on average over the full state of charge (SOC). Finally, it is shown how the battery model can be combined with a Kalman Filter and implemented on embedded systems using model-based design (MBD).
Download slides or view talk on ETH Video Portal