Model Predictive Control Methods for Three-Level Sparse Neutral Point Clamped Inverter

Abstract

Three-level sparse neutral point clamped inverter is employed in this paper due to low harmonic, fewer switching devices, and less dc-link capacitor voltage imbalance problems. Model predictive control (MPC) methods are proposed to fully exploit the advantages of this inverter while ensuring neutral point voltage balance and maintaining low computational complexity. First, the desired voltage vector reference is constructed to realize reference current tracking based on the model of three-level sparse neutral point clamped inverter. Second, the n-type or the p-type small voltage vectors are carefully tuned to minimize the neutral-point (NP) voltage unbalance. Last, the voltage vectors that are far from the desired voltage vector reference are excluded in the cost function so as to reduce the computational burden. Therefore, only 5 effective voltage vectors instead of totally 13 voltage vectors are required in the implementation of the proposed fast MPC algorithm, which saves computational time up to 61.5%. Moreover, it compensates the one cycle digital delay through predictions. The performance of the proposed method has been validated by the simulation and experimental results.