As the inverter station method for control of power supply to passive networks is usually through the outer voltage loop and inner current loop

there exist such shortcomings as a complicated control structure

four proportional-integral (PI) parameters to be adjusted

a slow response for load disturbance and poor voltage quality when supplying unbalanced or nonlinear load. Therefore

a transient mathematical model of modular multilevel converter based high-voltage direct current (MMC-HVDC) inverter side in stationary reference frame is developed. Then the discrete mathematical model is derived

and then model predictive direct voltage control (MPDVC) for the inverter station is designed. To reduce the computational burden of MPDVC

deadbeat model predictive direct voltage control (DBMPDVC) is presented

which is a combination of deadbeat control with model predictive control. A 21-level MMC-HVDC model is set up in PSCAD/EMTDC to validate the effectiveness of the proposed control method under a sudden increase in load

voltage rise

and supplying asymmetric three-phase load or nonlinear load

and furthermore a simulation comparison with the outer voltage loop and inner current loop control method is presented. Simulation results show that DBMPDVC can provide high voltage quality power for a passive network under various operation conditions and its control performance is obviously superior to outer voltage loop and inner current loop control method obviously.