Search

The state-of-charge (SOC) estimation approaches based on the pack model can hardly provide precise estimation due to cell difference, while the approaches based on each cell cost high computation resource, which are not suitable for real-time application. The estimation error can be further enlarged without considering balancing current. The SOC estimation approach of the battery pack considering balancing current is proposed, which dynamically searches for the cell with maximum or minimum voltage, and it only needs to calculate the selected cell in every estimation cycle. Compared to the approaches based on the pack model or each single cell, this approach can achieve precise pack SOC and cost less calculation time and resource. It has been verified with ten series-connected 200 Ah Li(NiCoMn)O2 batteries. The SOC estimation error is limited to 0.3% during the charging process, and a reduction of 2.5% is achieved compared to an error of 2.8% based on the pack model. A reduction of 1% is achieved compared to an error of 1.5% based on the pack model during the discharging process. Compared to an error of 1.7% without considering balancing current, a reduction of 1.4% is achieved during the charging process. [ABSTRACT FROM PUBLISHER]

Lithium-ion (Li-ion) batteries are nowadays the best tradeoff between performance, cost, and lifetime to store electric power and energy effectively. The reliable determination of the batteries’ instantaneous power capability and energy content is mandatory for many mobile high-energy and power-consuming applications. Herein, the battery impedance and state of charge (\SOC) are the relevant values. In energy storage systems, these values are to be considered for each battery cell individually, due to arising limitations caused by cell-specific variations. Simple algorithms to determine the battery system’s impedance and \SOC are presented including parameter and state-estimation techniques. Furthermore, methods are derived making the impedance and \SOC determination possible for a large number of particular cells in a battery system. The applicability of the demonstrated algorithms for battery control unit implementation is proved incorporating data achieved from load scenarios applied to a cell stack Li-ion battery module. Cell impedance and \SOC variations could be detected precisely. Moreover, a combination of impedance and \SOC spread was identified during typical battery operation. [ABSTRACT FROM PUBLISHER]