Model Predictive Torque Control for Dual Three-Phase PMSMs with Simplified Deadbeat Solution and Discrete Space-Vector Modulation

Unprecise voltage vectors applied in conventional model predictive control (MPC) would cause additional ripples in electromagnetic torque. To eliminate the problem, this article proposes an improved model predictive torque control (MPTC) based on deadbeat solution and discrete space-vector modulation (DSVM) for dual three-phase permanent magnet synchronous machines (PMSMs). First, deadbeat-direct torque and flux control (DB-DTFC) algorithm is applied and simplified, so that the computational burden can be reduced. Second, the virtual voltage vectors (VVVs) are adopted in dual three-phase voltage vector space to reduce voltage harmonics. Also, the DSVM is further proposed in this VVVs based vector space to generate more voltage vectors. After that, the simplified DB-DTFC and the DSVM scheme are artfully combined to select suitable voltage vector candidates for the MPTC, and the best one is then chosen to control the dual three-phase PMSMs. Finally, both simulation and experimental results are given to verify the effectiveness of the proposed method.

I. Introduction

Multiphase electric machines have drawn much attention from both academia and industry recently due to the advantages of lower rate power per phase, higher reliability and more control degrees of freedom, when compared with the three-phase counterparts [1]–[3]. Among the multiphase machines, the dual three-phase permanent magnet synchronous machines (PMSMs), also known as asymmetrical six-phase PMSMs, are with great application potential, because they not only have the merits of multiphase machines but also inherit the widely-used three-phase technology [4], [5]. Besides, they possess the feature of permanent magnet machines like high efficiency and high power density. Thus, dual three-phase PMSMs have been widely analyzed, like in the field of electric vehicles [6] or energy storage system [7].