The IPOS-DAB converter, which takes dual-active-bridge converters (DAB) as a sub-module, and with input parallel output series (IPOS), is the key hub of a full-DC system realizing wind power DC convergence and delivery. The component parameters of an IPOS-DAB converter in actual engineering application are different, and the output power of different DAB modules when parallel/series connection is unbalanced results in DAB module output voltage and current imbalance. This will increase the sub-module damage rate, greatly reducing system reliability. Conventional voltage equalization control suffers from low reliability and insufficient dynamic performance, and therefore cannot meet the control requirements during external dynamic disturbances. To address these problems, this paper uses the advantages of a model predictive control (MPC) method and adopts it for voltage equalization control of an IPOS-DAB converter. First, the prediction model is established from the IPOS-DAB converter. Then, the control objective function and the overall control strategy are designed. Finally, the system model is constructed in Matlab/Simulink. Compared with traditional voltage equalization control in different working conditions, the proposed control strategy has a better voltage equalization control effect.