Direct model-based predictive control scheme without cost function for voltage source inverters with reduced common-mode voltage

Abstract

This paper presents a direct model-based predictive control scheme for voltage source inverters (VSIs) with reduced common-mode voltages (CMVs). The developed method directly finds optimal vectors without using repetitive calculation of a cost function. To adjust output currents with the CMVs in the range of -V-dc/6 to + V-dc/6, the developed method uses voltage vectors, as finite control resources, excluding zero voltage vectors which produce the CMVs in the VSI within +/- V-dc/2. In a modelbased predictive control (MPC), not using zero voltage vectors increases the output current ripples and the current errors. To alleviate these problems, the developed method uses two non-zero voltage vectors in one sampling step. In addition, the voltage vectors scheduled to be used are directly selected at every sampling step once the developed method calculates the future reference voltage vector, saving the efforts of repeatedly calculating the cost function. And the two non-zero voltage vectors are optimally allocated to make the output current approach the reference current as close as possible. Thus, low CMV, rapid currentfollowing capability and sufficient output current ripple performance are attained by the developed method. The results of a simulation and an experiment verify the effectiveness of the developed method.