Multi-step predictive control will significantly reduce the switching frequency of a power conversion system at the same current quality, which is suitable for high power offshore wind turbine systems. However, the computational burden increases exponentially with the prediction horizon, making it difficult to apply in practical. In order to solve the above problems, a multi-step predictive control method based on ant colony optimization is proposed. The proposed method maintains good control performance and reduces the computational burden by 90% compared to traditional enumeration strategies. The innovations of the proposed method are: An ant colony optimization framework for back-to-back converter PMSG wind turbine systems with multi-step MPC is constructed, and a pheromone update mechanism is designed to fit the system. The relevant parameters design is also introduced.