Battery Health Monitoring and Ageing Prediction

Battery management systems face an increasingly complex plethora of additional tasks that needs to be solved as battery technology evolves and a large amount of data becomes available from fleet operational data. Some of these challenges are:

  • Ageing prediction
  • Health Monitoring
  • Ageing-aware charging and power split strategies.
  • Thermal Management

In order to tackle these, accurate modelling of the battery pack and correct state of charge estimation is required, with a varying degree of complexity depending on battery chemistry, temperature and so on. Additionally, for hybrid electric vehicles, power split strategies that considers both battery and fuel cell degradation and operating limits are vital for the extension of the useful lifetime of such vehicles.

Blue car with three symbol circles above

Ageing Prediction Modelling

The goal here is to create a model which predicts the battery degradation given a certain usage pattern. In order to create these models, a large amount of ageing data is required, where data from different batteries tested in parallel, with several different load profiles, is gathered. Different methods on how to create these models are available in literature, from electrochemical approaches to machine learning solutions. Also, given the high monetary cost in running these tests, it is relevant to design them in a way such that the model is obtained with a good quality and testing is kept to a minimum.

Health Monitoring

In order to take the necessary corrective maintenance steps and replace the battery when needed, it is necessary to have a methodology that enables the tracking of the battery health over-time, in a real time fashion. Since usually these methods require an accurate state-of-charge signal, adequate current-voltage models for the batteries are required, especially for chemistries with a flat OCV curve.  

Power Split and Thermal Management

For hybrid fuel-cell electric vehicles, where the motive power is provided by either a fuel cell or the battery, optimal power allocation of vital importance. A trade-off strategy was developed in order to operate both the battery and the fuel cell in a way that the efficiency is maximized, while respecting operational constraints on both sides, together with a cooling strategy that keeps the battery temperatures in a desired range.

Publications

Hametner, Christoph, Stefan Jakubek, and Wenzel Prochazka. "Data-driven design of a cascaded observer for battery state of health estimation, opens an external URL in a new window." IEEE Transactions on industry applications 54, no. 6 (2018): 6258-6266.

Hametner, Christoph, Bernhard Brunnsteiner, and Wenzel Prochazka. "Identification of capacity-loss prediction models of lithium-ion batteries, opens an external URL in a new window." In Proceedings of 7th International Symposium on Energy. 2017.

de Oliveira Jr, Jose Genario, Vipul Dhingra, and Christoph Hametner. "Feature Extraction, Ageing Modelling and Information Analysis of a Large-Scale Battery Ageing Experiment, opens an external URL in a new window." Energies 14, no. 17 (2021): 5295.

Nonlinear system identification based on nonuniformly sampled data for battery state of health estimation (NEW paper)

Lithium-ion Cell Ageing Prediction with Automated Feature Extraction (NEW paper)

Predictive Battery Cooling in Heavy-Duty Fuel Cell Electric Vehicles (NEW paper)

Adaptive Energy Management Strategy to Avoid Battery Temperature Peaks in Fuel Cell Electric Trucks (NEW paper)

Impact of Energy Management Strategy Calibration on Component Degradation and Fuel Economy of Heavy-Duty Fuel Cell Vehicles (NEW paper)

Research Projects at our Institute

Contact

Associate Prof. Dipl.-Ing. Dr.techn. Christoph Hametner

Send email to Christoph Hametner