Online update of battery modeling parameters and state of charge for parallel battery systems with incomplete current measurement

Accurate Parameter identification are essential for controlling the operation performance of battery systems. In practical battery systems with parallel-connected cells or modules, we commonly have incomplete current measurements of the uneven current distribution among parallel branches caused by inconsistent battery states or parameter values. Consequently, most battery parameter identification or SoC estimation methods proposed in recent literature struggle to provide accurate results due to missing some necessary measurement data. To address this challenge, this work proposes a method for battery parameter identification and SoC estimation on the fly. Based on an accurate and fast OCV prediction method we development previously and preliminary battery parameter values identified offline, initial battery states, such as the diffusion voltage in the battery model, are identified even with dynamic operating currents. Then, the missing current distribution among parallel battery branches are simulated based on the battery system model to facilitate the online battery parameter update. Iterative search can be performed to further improving the accuracy. This approach does not only bridge the gap caused by the incomplete measurement data but also features high estimation accuracy and computational efficiency, making it suitable for various practical battery-powered applications.