The significance of cell balancing is to balance the SOC and voltage between the cells when they are fully charged or discharged [Omariba, 2019]. The balancing algorithms detect and control the cell imbalances, classified as SOC, OCV, and terminal voltage based [Ma, 2018]. Using SOC or OCV provides better SOC estimation; however, the result depends on the SOC estimation algorithm's accuracy [Lin, 2020]. It is not an efficient approach for batteries with SOC curves versus flat OCV (e.g., LiFePO4). The balancing method using data-driven approach SOC estimation showing significant results but needs in-depth domain knowledge and large quanta of data [How, 2019]. The least complexity implementation approach based on voltage are extensively used in E-mobility systems [Tavakoli, 2020]. In BEV, at the final stage of the charging process in balancing is frequently done using voltage-based methods [Plett, 2004]. Final voltage algorithm balances in less time hence needs higher balancing current level and it works at the end of charging [Chen, 2018].The module level balancing [Han, 2019] is hierarchical in with respect to pack, module, and cell level, which is effective for large-scale energy segments. [Han, 2019] proposed a dynamic reconfiguration-based emerging balancing technique, capable of matching the user behaviour and requirements by changing the battery interconnection. An optimal control approach [Ouyang, 2020], [Docimo, 2019] accomplishes the key cell balancing objectives such as reduction in energy dissipation, shortening balancing time, increasing efficiency and capacity, and reducing the temperature raise. BMS with passive balancing for Li-ion batteries replacing lead-acid batteries in an electric kart is designed by [Vitols, 2014] to increase the pack's current capability and energy/mass ratio. [Amin, 2017] developed a passive balancer for 15 series connected cells by using internal MOSFET resistor for balancing, reduction in BMS hardware size, and improvement in balancing current to 3.07A, hence reduction in balancing time is seen. However, for the higher imbalance scenarios, it takes more time to balance. The dissipative shunt resistor is replaced with MOSFET [Xu, 2019] to enhance the performance while reducing the system cost. The amplifier mode of MOSFET is triggered during the imbalance and the excess energy of the highly charged cell is dissipated through it.
It is possible to regulate the balancing current up to 1.2 A. The frequency of the balancing switch ON/OFF from 1,150 to 2 cycles and the SOC variance to 0.157 are both greatly reduced by the outlier detection-based balancing method mentioned in [Piao, 2015]. The algorithm can accurately forecast the abnormal cell, however some alterations in the voltage cut-off value have been noted. Despite the high circuit cost and complexity, [Schmid, 2017] investigated a passive balancing architecture that balances the cells electrochemically without the addition of any electrical devices. According to Valchev (2018), balancing at various charging currents (such as 1C, 0.5C, etc.) reduces the balancing time by 20%.
