A Deadbeat Control Algorithm Based on Adaptive Prediction Theory for Active Power Filter

Deadbeat control as a discrete system control method is widely used in active filter control system. But without the prediction deadbeat control is actually a difference a beat or two beats control, for this, the paper proposes a prediction based on adaptive line enhancer for active filter deadbeat control, accurate predicts system state after two beats with the historical datas. Theoretical derivation proves the adaptive prediction method' s robustness is optimal in the sense of H∞, experiments demonstrate that the proposed method is correct and feasible.