In this work, a battery active grouping equalization control strategy based on model predictive control (MPC) was proposed, which can promote cell consistency, equalization speed and energy loss during the battery equalization process. The dynamic group equalization topology based on reconfigurable circuits can achieve dynamic grouping. Using a battery state observation estimator and the MPC controller, multiple non-adjacent cells can realize simultaneous equalization in a single equalization process. An algorithm is designed to determine the optimal energy transfer direction and the optimal equalization current. The objective function of this algorithm incorporates weight coefficients that represent the relative importance of equalization time and energy loss. Simulation tests are conducted to evaluate the battery pack state-of-charge (SOC) root mean square, average temperature, and equalization time under various weight coefficients. Compared with two other traditional equalization control strategies, the proposed strategy reduces the equalization time by 43.93%, decreases the battery pack SOC variance by 50.18%, and improves the energy transfer efficiency by 0.59%.