Equalization strategy of lithium-ion battery packs under two-level structure: An adaptive model predictive control approach

Inconsistency is an inevitable problem of electric vehicle battery packs, which will lead to system performance degradation and increase safety risks. In this article, we propose a two-level equilibrium topology structure for inter-group and intra-group dynamics. The intra-group equilibrium topology is based on Buck–Boost converters, which balance any individual battery within the group by flexibly switching modes. The inter-group topology utilizes a Cuk circuit, which allows for energy transfer between every two adjacent battery groups through a dual-layer switch. Compared with the traditional Buck–Boost topologies, the proposed topology has the advantages of simple modular structure and rapid energy transfer speeds. An adaptive model predictive control (AMPC) balancing strategy is proposed to update and adjust the equalizer current. Compared with the traditional algorithms, this algorithm adopts a virtual output compensation (VOC) method to reduce the impact of model linearization errors, thereby improving system robustness and optimizing control performance. Finally, 12 batteries are randomly selected for different initial state of charge (SOC), which are divided into three groups. By conducting static, charging/discharging and consistency tests, it shows that the proposed strategy possesses excellent balance performance.