For the open winding permanent magnet synchronous motor (OW-PMSM) with leg open-circuit faults, a novel zero-sequence current (ZSC) suppression method with asymmetric zero-sequence voltage (ZSV) injection is proposed in this article. To realize the fault-tolerant control (FTC) in leg open-circuit fault conditions, the leg-sharing method is developed, where the output terminal of the faulty leg is connected to that of the remaining equivalent leg. However, the ZSC increases in the FTC mode, which causes severe torque ripple and system performance degradation. If the traditional ZSC suppression method is still adopted, the suppression performance is decreased. Besides, an undesired voltage vector is introduced, resulting in current distortion. To solve these problems, a novel ZSC suppression strategy is put forward. By asymmetrically adjusting the voltage output of the remaining legs, the ZSC of OW-PMSM in FTC mode can be effectively suppressed. Moreover, the proposed strategy is suitable for both the single-leg and double-leg FTC modes, where the ZSC suppression performance is almost the same as that of the traditional method in normal operation mode. Furthermore, experiments are performed on an OW-PMSM to verify the effectiveness of the proposed ZSC suppression scheme.

In this article, a novel phase current reconstruction strategy for a five-leg dual-inverter (FL-DI) fed open-winding permanent magnet synchronous motor (OW-PMSM) drive is put forward. Benefiting from the doubling of the power legs, the OW-PMSM drive is endowed with outstanding fault-tolerant ability. When an open-circuit fault occurs, the OW-PMSM automatically turns to the postfault two-phase operation. By reconfiguring the wiring method, the fault-tolerant FL-DI topology is obtained, and the postfault three-phase operation is realized. However, the current flow path changes and measurement vectors become unavailable in postfault operation processes, making the existing current reconstruction schemes unusable. To prevent the system collapse caused by the current reconstruction failure, a novel current reconstruction strategy is proposed, which can be applied in the case of the open-circuit fault in any leg. In addition, the proposed current reconstruction strategy is compatible with the specific zero-sequence current suppression for the FI-DI fed OW-PMSM drive, which ensures better system performance in the postfault operation. The reconstructed current is of high accuracy, where the relative reconstruction error is 7.8% at low load and 5.3% at heavy load. Finally, experimental results are presented to verify the feasibility and effectiveness of the proposed strategy.

Aiming to enhance the robustness and transient performance under model parameter mismatch, this article proposes a self-tuning incremental model-based predictive current control (STIM-PCC) strategy for permanent-magnet synchronous motor drives. In conventional incremental predictive current control scheme, although parameters such as permanent magnetic flux linkage and stator resistance are not required, stator inductance is still necessary for the predictive model. When the stator inductance mismatch occurs, prediction error is inevitable, leading to weakened robustness and deteriorated performance. To solve this issue, a novel STIM-PCC strategy is proposed, where the prediction error that indicates the inductance mismatch is adopted to tune the nominal inductance. In this way, the incremental model is updated adaptively and the precise predictive control can be achieved. Moreover, to enhance the tuning accuracy in the overmodulation region, a voltage constraint compensation method is put forward, which can effectively reduce the current drop during dynamic process. Compared to conventional scheme, the parameter robustness is significantly strengthened, where the inductance mismatch can be detected and corrected in time. Besides, not only the current fluctuation is reduced, but the settling time is shortened, thus greatly improving the transient performance. Experiments are carried out to validate the effectiveness of the proposed scheme.

In a hybrid-inverter open-winding permanent magnet synchronous motor (OW-PMSM) drive, the motor and the floating capacitor are controlled simultaneously. Besides, since the compensatory inverter barely provides active power under steady-state operation, the unity power factor control (UPFC) of the main inverter (MI) is also required. However, the conventional model-based predictive control of the hybrid-inverter OW-PMSM is highly sensitive to the accuracy of parameters, including the floating capacitance and stator inductance. Serious current distortions and system oscillations appear when the model parameters are mismatched with the actual ones. In addition, unmodeled parts such as power switch voltage drops and disturbances also degrade the system performance. Hence, a linear disturbance observer (LDO) based model-free predictive control (MFPC) is put forward. By using ultralocal models to substitute the mathematical models of the floating capacitor and OW-PMSM, the proposed scheme maintains satisfactory performance when system parameters are mismatched. Moreover, because system disturbances are obtained by LDO, the tracking errors can be eliminated even when parameters of ultralocal model are less appropriate. Thus, high precision of UPFC for the MI is guaranteed. Finally, experiments are carried out on a 500 W hybrid-inverter OW-PMSM drive, proving the superiority and effectiveness of the proposed MFPC strategy.

In this article, a novel current error calibration scheme considering noncentral sampling is proposed for an open winding permanent magnet synchronous motor (OW-PMSM) drive with a single current sensor. To realize the current reconstruction using a single current sensor, the durations of measurement vectors need to last long enough. Besides, the pulsewidth modulation (PWM) of the dual inverters is complicated, where the switching signals of two inverters need to be calculated simultaneously. Hence, the modified PWM strategy is developed to simplify the PWM procedure and extend the duration of measurement vectors. Nevertheless, the sampling error exists, because the bus current samplings take place sequentially at the measurement vectors instead of simultaneously at the central point. This will lead to a significant error in the reconstructed current, degrading the operation performance. Therefore, an online current calibration scheme is proposed to compensate for the current sampling error, which requires neither observer nor offline measurement to obtain the back electromotive force (back EMF). With the proposed calibration strategy, the precision of the reconstructed current can be significantly improved. Experiments are performed on an OW-PMSM drive to verify the effectiveness of the proposed strategy.

This article proposes a phase current reconstruction strategy for open-winding synchronous reluctance motor (OW-SynRM) drives with the dual-inverter topology and common dc bus. The traditional reconstruction strategy cannot be applied in OW motor drives because the dual-inverter modulation and zero-sequence current suppression strategies complicate the vector combinations and shorten their durations. Besides, additional sampling is required because the sum of the three-phase current is not zero. To solve these issues, the edge alignment algorithm is proposed to extend the duration of certain vector combinations, and the phase-shift strategy for the third sampling is proposed to obtain extra current information. In the case of the reconstruction failure in immeasurable regions, the corresponding compensation strategy is investigated. With the proposed reconstruction strategy, the system cost can be significantly reduced and the reliability can be improved, where only one current sensor is required. Besides, the single sensor can be reused for power detection and protection for the whole system by selective installation. Moreover, the proposed strategy is compatible with both the 180° and 120° assignment schemes, increasing the flexibility of application. Finally, experiments are carried out on an OW-SynRM to verify the high accuracy and feasibility of the proposed strategy.