Optimal design for efficiency enhanced hierarchical cell-to-cell equalization systems based on centralized model predictive control

Abstract

Effective topology is particularly significant for enabling high-efficiency hierarchical cell-to-cell (CC) equalization systems for large-scale batteries. This paper proposes an efficiency-enhanced hierarchical battery equalization systems (BESs) by formulating a topology graph-balancing connectivity evaluation method. As a performance evaluation index, connectivity takes both balance time and energy loss into consideration, which can be measured by the second small eigenvalue of the Laplacian matrix of the topology graph. Then, a centralized model predictive control (MPC) balancing strategy is formulated and integrated with hierarchical topologies to tackle optimal residual available energy (RAE) balancing problems. Making use of simulation tests of hierarchical BESs with different configurations, results show that the balancing performance and measured connectivity are consistent, and the balancing connectivity gradually rises and then decreases with an increasing number of module groupings. The optimal hierarchical BES with 96-series cells can achieve a low balancing time of 229 s and an energy loss of 0.67%. The balancing time and energy loss of centralized MPC can be reduced by 80.28% and 34.02%, respectively, relative to decentralized MPC balancing. For practical implementation consideration, dynamic balancing simulation and hardware-in-the-loop (HIL) experiments are conducted, and the experimental results are well matched with the simulation.