2026/07/01 by Chinni, Chinni Saiteja, Arnav Dutta +3
Engineering · #Advanced Battery Technologies Research #Advanced battery technologies research #Advancements in Battery Materials
paper · pdf · doi:10.1016/j.compchemeng.2026.109816
Al-air batteries have the advantage of giving longer mileage to electric vehicles. They have the advantage of not needing the charging-discharging process since they use water to generate power. Each Al-air cell can provide voltage of 2.7V and specific capacity of 8100 Wh/kg theoretically and many such cells can be stacked to form a battery that can power an electric vehicle (EV). Since many cells are being connected together, cell balancing is required. Active cell balancing is a crucial aspect of battery management systems (BMS) for optimizing the performance and lifespan of battery packs such as Li-ion, Al-air batteries etc. This paper proposes an innovative active cell balancing algorithm utilizing a multi layered neural network based adaptive neuro-fuzzy inference system (MLNN-ANFIS) method tailored specifically for Al-air batteries. The MLNN-ANFIS-based algorithm intelligently manages energy transfer among battery cells based on state of charge (SoC) differentials, aiming to mitigate cell imbalances and enhance overall battery pack efficiency. Through comprehensive simulations and experimental validations, the proposed algorithm's performance is evaluated and compared with conventional active cell balancing techniques. Results demonstrate that proposed ANFIS algorithm gives convergence in SoC and is 2 times faster than other neural network and fuzzy based active balancing algorithm, thus minimizing energy loss in Al-air battery systems. The algorithm brings about balancing for real-time laboratory assembled Al-air battery pack as well.