Abstract：To address the degradation of system robustness when applying traditional control strategies to

phase-shift full-bridge （PSFB） converters under complex operating conditions，a receding horizon optimal

predictive control based on adaptive super-twisting sliding mode observer was proposed. First，an ultra-local current equivalent model was introduced to effectively simplify the system modeling complexity. Next，to further enhance the smoothness of the sliding mode surface transition in the super-twisting sliding mode observer and improve its performance under nonlinear disturbances，an adaptive super-twisting sliding mode observer was designed based on the expected dynamic variation of current errors from the ultra-local model. The observer´s gain matrix coefficients were adaptively adjusted by real-time predictions of the inductor current error trend using an online least squares method. This approach enabled real-time estimation and compensation of lumped model disturbances，thus improving the system´s ability to adapt to nonlinear disturbances. Finally，experimental validation of the proposed algorithm demonstrates that it exhibits superior performance in terms of control accuracy and robustness.

        Key words：phase-shift full-bridge（PSFB）；adaptive super-twisting sliding mode observer；receding horizon

optimal predictive contro（l RHOPC）

     Format Citation:薛瑞斌，曾甲辰，黄东晓,等. 移相全桥变换器滚动时域优化预测控制策略[J].电气传动,2026,56(07):10-15. XUE Ruibin，ZENG Jiachen，HUANG Dongxiao,et al. Receding horizon optimal predictive control strategy for phase-shift full-bridge converter [J].Electric Drive, 2026,56(07):10-15