Bridge Arms Current Tracking Predictive Control of Modular Multilevel Converter for HVDC System

In order to suppress circulation current, a predictive control method based on error model of system is proposed. Firstly, we set up the discrete model of the controlled system by Eulor´s formula, and then derive the tracking error discrete model, finally, the desired AC current is tracked completely by the actual current. The merits of the error model predictive control is that the error model eliminates the repeat errors in the time index and the predictive control and Kalman filter remove the random error. The calculated amount is too heavy to be implemented by the computer in the model predictive control for modular multilevel converters when the numbers of the sub-modular is high, a explicit optimal grouping method is bring forward according to the mean inequality, and reduce the computer burden. The simulation result show the proposed control strategy is of feasibility and superiority.