Predictive control of a series-connected two five-phase synchronous magnet permanent machines

Abstract

The work presented in this paper focuses on studying the concept of two five-phase permanent magnet synchronous machines connected in series, powered by a single five-arm converter, in order to considerably minimize circulating currents. Firstly, we modeled a system composed of two identical five-phase synchronous machines, whose stator windings are connected in series. This modeling aims to demonstrate that the control of each machine in the group can be independent of the other, even though both machines are powered by a single five-phase converter. Then, we carried out the predictive control of the two five-phase machines connected in series. This control is based on the prediction of future switching states, which are managed via the discrete model of the system. The optimal switching state is obtained according to a predetermined cost function. To demonstrate the features of the presented control algorithm, the results of the simulation conducted using MATLAB/Simulink are presented and discussed to evaluate the performance and efficiency of the independent control of the two five-phase synchronous machines connected in series.