Abstract: Model predictive control has the advantages of intuitive concept, fast response and easy handling of nonlinear constraints, and is widely used in multiphase motor control systems. Most of the existing model predictive controls use virtual voltage vectors to reduce harmonic losses. However, the static nature of virtual voltage vectors leads to a lack of flexibility in their combination, which is suboptimal in nature. Considering the limitations of virtual voltage vectors, from the perspective of vector dynamic selection, a dynamic virtual voltage vector model predictive current control(MPCC) strategy is proposed, which generates dynamic virtual voltage vectors online during the current prediction process, through different stages, the optimal voltage vector required for the synthesis of the virtual voltage vector and the optimal voltage vector action time are determined step by step. This strategy overcomes the static characteristics of virtual voltage vectors to achieve effective suppression of harmonic currents. Finally, the experimental effects of different control strategies are compared, and the effectiveness and feasibility of the proposed dynamic virtual voltage vector MPCC strategy are verified.