Search

This paper presents a novel model predictive control (MPC) for the T-type single-phase three-level inverters. The cumbersome procedure in tuning of weighting factors is eliminated so that the implementation of the proposed MPC becomes easy. The constant switching frequency is achieved during the control implementation in order to facilitate the filter design. In order to optimize the distribution of the output current harmonics, multiple voltage vectors are employed in each control cycle with their application times setting inversely proportion to the respective cost function. The redundant small voltage vectors are utilized to balance the neutral point voltage of the inverter by directly regulating the upper and lower dc-link capacitors voltages. Finally, the proposed MPC algorithm is experimentally evaluated and compared with other two unfixed switching frequency MPC algorithms in terms of the steady state, transient performance, and parameter sensitivity. [ABSTRACT FROM AUTHOR]

• The proposed method is suitable for more electric aircraft grids. • The proposed method has fast dynamics. • The structure of the proposed method is simple. Different from utility power systems, the grids of more electric aircraft (MEA) are frequency-variable, and the frequency change range is from 360 Hz to 800 Hz. Most grid synchronization methods are designed for constant-frequency grids whose frequency varies in a small range around 50 Hz or 60 Hz. Therefore, the dynamic performance of these methods decreases considerably in MEA grids. In addition, the frequency in MEA is higher than that in utility power systems. Grid voltage parameters are thus required to exhibit a fast dynamic response. To resolve this problem, this paper proposes an algebraic method to estimate the positive- and negative-sequence components (PNSC) of grid voltage accurately. By analyzing the relationship between grid voltage and its time delay signals, the linear regression model is established. Then, grid frequency, PNSC are obtained with an algebraic method. The proposed method relies on algebraic operations only and thus exhibits a fast dynamic response. Experiment results verify the effectiveness of the proposed method. [ABSTRACT FROM AUTHOR]