Model based development of mild hybrid electrical battery system

Presented by Andreas Eriksson and Mohanapriya Anandaguru, MPEPO


In this thesis, a battery energy storage model is developed for a 48 V Mild Hybrid system in MATLAB/ Simulink environment. The battery energy storage model consists of an electrical model, a thermal model and a SOC estimator.  The electrical model is based on equivalent circuit models where both 0RC, 1RC and 2RC model parameters are extracted from measurements on individual cells. The thermal model is based on a 1D lumped parameter model including the thermal resistance from the battery pack to environment and the thermal mass of the battery pack. The SOC estimator are based on current integration. The battery model is evaluated during different driving cycles and the electrical models are compared in terms of accuracy. Finally the battery energy storage model is implemented in a simplified 48 V mild hybrid systems together with an ISGM model and is evaluated in terms of electrical and thermal behaviour.

The electro-thermal battery model shows good accuracy for a WLTP driving cycle, with a voltage RMSE value around 1% and a maximum error of 6% which occurred at low temperatures. When comparing the 2RC and 1RC electrical models, it shows that they have similar performance. The thermal model shows an RMSE of 1-2 °C and a maximum temperature error of 5 °C for the WLTP driving cycle. The SOC estimator shows a RMSE of 1% and a maximum error of 2%.

Supervisor: Stefan Skoog (Chalmers University of Technology), Linas Brusokas and Torbjörn Larsson (Volvo Car Corporation)
Examiner: Torbjörn Thiringer
Category Student project presentation
Location: Fredrik Lamm's room, Hörsalsvägen 11, Electric Power Engineering
Starts: 14 June, 2018, 09:30
Ends: 14 June, 2018, 10:30

Published: Thu 31 May 2018.