Battery Pack Modeling and Performance Assessment for Electric Vehicle Applications
Examensarbete för masterexamen
Automotive engineering (MPAUT), MSc
Large traction battery packs are made up of several number of cells connected in series and parallel to fulfill vehicle demanded voltage and capacity. For many years, the operation of the battery packs has been overestimated by only scaling up battery cell performance to mimic pack behavior. The inconsistencies in the current, State of Charge (SOC) and temperature of the individual cells/modules originating from the battery pack electrical topology and thermal circuit design, negatively influence the pack overall performance. In the current thesis work, lumped battery pack is modified to account for cell-to-cell variations. All simulations have been performed in Siemens Amesim, a commercial software for the modeling and analysis of multidomain systems. The inhomogeneities within the battery pack are modeled by considering interconnection resistance between parallel branches. The input current to the models is given either as a constant value or through a current profile obtained from literature survey and the software tutorial. Both the lumped pack and cell-to-cell discretized models (C2C) are compared in terms of model predictability. The results show that with C2C model the variations between the cells can be observed and quantified. The C2C model discharges 11% faster compared to the lumped pack model when Worldwide harmonized Light-duty vehicles Test Procedure(WLTP) current profile is inputted. The maximum temperature attained by C2C model is 12 0C greater than the lumped pack model. It can be concluded that when lumped pack modelling approach is employed the vehicle achievable range is overestimated and the maximum temperature is underestimated.
battery pack modelling , equivalent circuit method , cell-to-cell variations , electric vehicle