Predictive Deadbeat Average Model Control for Power Conversion System With LCL Filter

Abstract

The inherent resonance issue of LCL filters in the power conversion system(PCS) will significantly degrade the system’s stability. Researches indicate that decreasing the time delay contributes to the system’s stability under converter-side current feedback(CCF) control. To improve the system’s stability, this paper is devoted to alleviating the impact of time delay on the system’s stability by the combination of time-delay compensation and time-delay reduction. Since the effectiveness of time-delay compensation provided by predictive control depends on the accuracy of the control model, a deadbeat average model (DBAM) for the CCF loop is established based on the use of the oversampling technique. The proposed control model, which ingeniously transforms conventional single-point sampling control into multi-point average control, improves the accuracy of the system description. Furthermore, through the introduction of predictive control combined with the double-updated PWM method, a predictive deadbeat average model (PDBAM) control method is proposed for the PCS. The proposed control method minimizes the impact of time delay on CCF loop, which enhances the system’s stability and overcomes the resonance issue of LCL filter. Besides, the stable region of the PDBAM control is deduced. Simulation and experimental results demonstrate the effectiveness of the PDBAM control.

Keywords

• LCL filter
• power conversion system
• predictive control
• system stability