Modeling of LFP cells with C-rate and temperature dependence for state-of-charge estimation

Beskrivelse

This objective of this study was to develop a second-order RC (2RC) equivalent circuit
model (ECM) and a robust state-of-charge (SoC) estimation algorithm for Lithium
Iron Phosphate (LFP) cells, while accounting for temperature and discharge C-rate dependencies.
The work is motivated by the need for precise battery modeling to solve
the problem of LFP’s inherently flat Voltage-SoC relationship and significant parameter
variations under dynamic operating conditions. The research methodology involved
characterizing two identical 1.8 Ah LFP cells through Hybrid Pulse Power Characterization
(HPPC) and Open-Circuit Voltage (OCV) measurements across temperatures of
0°C, 25°C, 45°C and discharge C-rates of 0.5 C, 1 C, 2 C, which yielded the model’s
electrical parameters. The 2RC equivalent circuit model which was constructed using
these parameters was subsequently validated against independent data. An Extended
Kalman Filter (EKF) was then implemented using a combination of coulomb counting
and model-based correction to form a real-time SoC estimation algorithm. The accuracy
of the estimation was assessed using Root Mean Square Error metrics and results showed
EKF accuracy improved with increasing temperature and C-rate, but struggled in cold
conditions. Cell-to-cell variations were also observed across the experimental results.
Overall, this work contributes a battery model which captures Temperature and C-rate
fluctuations alongside an effective SoC estimation strategy. This lays a crucial foundation
for future research into creating better SoC estimation for LFP cells when used in its
numerous applications which includes electric vehicles and energy storage applications.