Search

Multiphase machines are increasingly used in research and industry applications due to their inherent advantages. Stator current control is a common strategy for this type of systems. The most important issue it must face is regulation of currents in the torque producing plane and the harmonic plane. For this task, finite control set model predictive control (FCS-MPC) constitutes an interesting alternative to methods using modulation. However, the implementation of FCS-MPC is characterized by a high computational demand, limiting the sampling frequency. This work proposes a predictive algorithm that needs less computation time. As a result, the sampling period can be reduced while producing predictive control. This brings about several benefits resulting from improved current tracking. The proposed method avoids the combinatorial optimization phase of standard FCS-MPC, which is the most time-consuming part. The algorithm is based on physical insights obtained from the application of FCS-MPC to multiphase drives leading to the concept of evolutionary gaps regions. The experimental results for a five-phase motor demonstrate improved performance. Moreover, the method is flexible enough to balance the tradeoff appearing between the torque producing plane and the harmonic plane.

Finite state model predictive control (FSMPC) of multiphase drives can use an extra number of inverter configurations compared with the three-phase case. This, however, requires more computing power for the optimization phase. The application time of each selected voltage vector (VV) is then increased, which can result in higher harmonic content. Reducing the allowed VVs can speed up the computations, thus ameliorating the current tracking/regulation in different orthogonal subspaces. However, the flexibility offered by the reduced set of VVs is less than that of the full set. Furthermore, a lower sampling time can result in an increase in the switching frequency, especially for some speed-load combinations. This article proposes the use of a hybrid scheme where the set of allowed VVs is not fixed but rather selected on-line according to the actual speed and torque producing stator current which are computed by the outer loop. A five-phase induction machine (IM) is used as a test bed for the proposal, showing improved results with respect to the nonhybrid case.

Model predictive control (MPC) has been recently suggested as an interesting alternative for the regulation of multiphase electric drives because it easily exploits the inherent advantages of multiphase machines. However, the standard MPC applies a single switching state during the whole sampling period, inevitably leading to an undesired $xy$ voltage production. Consequently, its performance can be highly degraded when the stator leakage inductance is low. This shortcoming has been, however, mitigated in recent work with the implementation of virtual/synthetic voltage vectors (VVs) in MPC strategies. Their implementation reduces the phase current harmonic distortion since the average $xy$ voltage production becomes null. Nevertheless, VVs have a static nature because they are generally estimated offline, and this implies that the flux/torque regulation is suboptimal. Moreover, these static VVs also present some limitations from the point of view of the dc-link voltage exploitation. Based on these previous limitations, this article proposes the implementation of dynamic virtual voltage vectors (DVVs), where VVs are created online within the MPC strategy. This new concept provides an online optimization of the output voltage production depending on the operating point, resulting in an enhanced flux/torque regulation and a better use of the dc-link voltage. Experimental results have been employed to assess the goodness of the proposed MPC based on DVVs.

The inherent fault-tolerant capability of multiphase machines is highly appreciated, but it requires fault detection and localization together with a reconfiguration of the control scheme. When the multiphase machine is regulated using finite-control set model predictive control (MPC) strategies, the reconfiguration involves the use of different transformation matrices, cost functions, and current references for each of the multiple open-phase fault (OPF) scenarios. Aiming to simplify this procedure and add further robustness, this paper explores the possibility to achieve a natural fault-tolerant capability by maintaining the prefault control strategy after the fault occurrence. For this purpose, this paper first analyzes the two main reasons why MPC-regulated multiphase drives misbehave in the event of an OPF: the voltage vector shifting and the search for incompatible goals. In the next step, a version of the MPC that includes virtual-voltage vectors (VVs) is tested for the first time in postfault situation and it is compared to conventional MPC technique. Extensive experimental results reveal that, while MPC misbehaves in the event of an OPF, the VV-MPC provides a satisfactory ripple-free postfault performance. This finding has two significant implications for industrial applications: the postfault operation is highly simplified and, at the same time, the fault-tolerant multiphase drive becomes immune to fault detection errors and delays.