Model predictive control for the high-power grid-connecting PWM converters

Large-scale network of the renewable energy imposes increasing stringent requirements, especially with regard to harmonic distortion limits imposed by the grid code and the switching losses of the high-power PWM converter itself. This paper presents a new control strategy adopting the receding horizon policy, i. e., model predictive control (MPC), for a diode clamped three-level topology converter. With a simplification of the three-level SVPWM into a two-level SVPWM and taking the grid current tracking error, harmonic distortion limits, switching numbers per cycle and the deviation of the neutral point potential as components of the cost function, a relative good performance could be obtained with a high degree of robustness because of the utility of receding horizon policy.