An Improved Low-Complexity Model Predictive Direct Power Control With Reduced Power Ripples Under Unbalanced Grid Conditions

Abstract

This article proposes an improved low-complexity model predictive direct power control (LC-MPDPC) along with reduced power ripples. Different from those MPDPC methods under balanced grid voltage conditions, both the design and implication of MPDPC under unbalanced grid conditions are further studied. In the proposed MPDPC, to reduce the computational burden of the controller, an improved LC-MPDPC based on the extended reactive <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">pq</i> theory is proposed. Only one prediction is needed to select the next voltage vector. Besides, the defined cost function under balanced grid conditions cannot obtain the minimized value if the unbalanced grid conditions occur. Accordingly, a new power error is defined, which is described analytically in the optimal model. With the proposed LC-MPDPC, a negative conjugate of the new defined complex power is selected as the control variable, and an improved modulation for the LC-MPDPC is presented to ensure the best voltage and calculate the optimal duration under unbalanced grid conditions. Compared with prior MPDPC, the proposed LC-MPDPC obtains much lower power ripples and grid current total harmonic distortion, and further, it offers a fast dynamic response and robustness. The effectiveness of the proposed LC-MPDPC is evaluated and validated by the experimental results.