Enhanced Natural Fault-Tolerant Model Predictive Current Control in Nine-Phase Motor Drives Under Open-Phase Faults

For the open-phase fault (OPF) tolerant operation of multi-phase motors based on model predictive current control (MPCC), the fault diagnosis and control scheme reconfiguration are difficult to avoid, which leads to the increase of control complexity. A natural fault-tolerant (NFT) MPCC method based on virtual voltage vectors (V
$^{3}$
) can simplify this procedure and realize the fault-tolerant operation by using a unified control algorithm during pre- and post-faults. However, there is no effective control scheme to deal with the V
$^{3}$
shifting after the fault, which will lead to the control errors. To overcome this problem, an enhanced NFT-MPCC strategy is proposed in this paper. Firstly, this paper analyzes the causes of the control error in NFT, and elaborates the influence of V
$^{3}$
shifting on the current tracking performance. Then, a duty ratio optimization based on an PI regulator is proposed for the first time. After the fault occurs, the PI regulator automatically corrects the duty ratio calculation result to eliminate the control error caused by vector shifting, and the proposed method is versatile to different OPF scenarios. Compared with existing NFT-MPCC technologies, the experimental results show that the proposed method significantly suppresses the torque ripple in post-fault, and provides a satisfactory control performance before and after fault without the need of software reconfiguration.