Fault diagnosis is currently considered a crucial method to enhance the reliability of permanent magnet synchronous motor (PMSM) drives. In this article, an open-circuit fault (OCF) diagnosis method of a voltage source inverter-fed PMSM with model predictive control is proposed. The developed approach mainly consists of fault detection and localization. Besides, the normalized cost function variation is designed as a fault detection indicator according to the impact of OCFs on current signatures. It can not only monitor drive system conditions, but also discriminate fault categories. Afterward, the polarities of normalized αβ -current average values and phase angles of residual current vectors are utilized to locate the exact faulty switch in each fault type. Consequently, a total of 21 possible combinations of faulty switches can be correctly detected and located. Finally, the effectiveness in fault diagnoses and the robustness against operating point variations are verified by the experimental results.

To reduce the thrust force ripple of a complementary and modular linear flux-switching permanent magnet machine, a model predictive thrust force control (MPTFC) is developed and implemented based on the active voltage vectors selection (AVVS) and two-voltage-vector synthesis (TVVS). Firstly, the active VVs are screened and assigned to the specific sector, instead of traversing all the possible vectors, and consequently, significantly decrease prediction workload. Secondly, for the elimination of the weighting factor in the conventional MPTFC, the control object is transformed from the thrust force and stator flux to the stator flux only by using the load angle. Thirdly, to achieve the steady-state performance improvement, the TVVS approach is applied in each control period, where the optimal TVVS and its dwell time are simultaneously obtained by only one cost function. Both simulated and experimental results verify that the steady-state performance improvement can be realized by the developed MPTFC through comparing with normal hysteresis current control and conventional MPTFC strategies.

To address both single open-switch fault (OSF) and arbitrary two OSFs of a five-phase permanent magnet synchronous motor (PMSM) drive, a generalized fault-tolerant model predictive current control method is proposed in this article. Firstly, the postfault currents under OSFs are analyzed in the stationary frame based on the Fourier method. Then, to maintain torque output capacity and alleviate torque ripple, the optimal reference currents in fault scenarios are deduced with the principle of rotating magnetic motive force undisturbed. Further, the obtained phase currents are transformed into the rotating frame and applied in the cost function for the voltage vector optimization. Thus, the torque under post-fault operations is guaranteed and the torque ripples are suppressed. Finally, various experiments are conducted to validate the feasibility of the generalized fault-tolerant control method of five-phase PMSM drives.

Fault tolerant capability is an important advantage of multiphase motor drives. To enhance the fault-tolerant performance of a five-phase flux-switching permanent magnet (FSPM) motor drive under the open-circuit condition, a model predictive current control (MPCC) with the simplified control set (SCS) is proposed in this paper. On the principle of minimizing the harmonic voltage in x-y subspace, two zero switching states and the switching state, which generates a larger voltage vector in α-β subspace, are selected into the SCS. Accordingly, the number of current predictions has been reduced, resulting in the decrease of computational cost. Besides, the duty cycle optimization (DCO) approach is applied in conjunction with the SCS-based MPCC to improve the steady-state performance. Finally, the effectiveness of the proposed MPCC method for an open-circuit fault-tolerant five-phase FSPM motor drive is validated by experiments. Comparative experimental studies show that the MPCC strategy with SCS and DCO can improve both the transient and steady-state performances of the fault-tolerant drive.

In this letter, a hybrid model-based diagnosis method is proposed for open-switch faults in permanent magnet synchronous motor drives by combining the advantages of both model predictive control (MPC) and mixed logical dynamic (MLD) models. This idea is based on the receding optimization of the MPC and the estimating capability of the MLD. The predictive current model and the MLD model are established to predict and estimate phase currents, respectively. Besides, the logical vectors are theoretically investigated under various faulty conditions. Derived from the cost function, a diagnosis function is designed to evaluate the consistency between the actual and estimated currents. Consequently, the logical vector minimizing the diagnosis function is selected to locate faulty switches. Finally, experiments are conducted to verify the effectiveness of the proposed method.

To enhance the steady-state performance of the virtual-vector-based model predictive current control (VV-MPCC) for a five-phase flux-switching permanent magnet motor, an improved multi-vector-based MPCC (MV-MPCC) method with geometric solution is proposed and investigated in this paper. The desired voltage is employed to pre-select voltage vector candidates and provide a guidance for the geometric division of a sector. According to the link between the reference voltage and the division, the optimal voltage vectors are determined and their duty ratios are obtained by projection method. Compared to the conventional MV-MPCC method using derivative method, the proposed geometric MV-MPCC method employs the desired voltage to quickly determine the voltage vectors and the corresponding duty ratio, which not only alleviates computational burden, but also favors for more intuitive implementation. Finally, the effectiveness of the proposed method is verified by experiments.