多相永磁容错电机常应用于对驱动系统可靠性要求较高的航空、交通运输和军事等领域。开路故障是电机驱动系统中常见的电气故障，针对现有故障诊断方法需要多组变量识别开关管和绕组开路故障，且电机运行状态发生变化易导致判断出现误诊断的问题，本文提出一种改进的相电压平均值算法，该方法将电机驱动系统与电流型磁链观测器相结合，将诊断变量由故障复杂识别切换为一组变量识别，由阈值范围诊断切换为正负变量组和诊断。通过诊断结果即可区分不同开关管及绕组开路故障。搭建了六相永磁容错电机故障诊断硬件实验平台，实验结果验证了所提算法的有效性及优越性。

Multi-phase permanent magnet fault-tolerant motors are often used in aviation, transportation and military fields which require high reliability of drive systems. Open circuit fault is a common electrical fault in motor drive system. Aiming at the problems that the existing fault diagnosis methods need to identify the open circuit fault of switch tube and winding with multiple sets of variables, and the change of motor running state is easy to lead to the misdiagnosis of threshold value judgment, this paper proposes a new phase voltage average value algorithm, which combines the motor drive system with the current mode flux observer. and switches the diagnosis variables from fault complex identification to a group of variables identification, and from threshold range diagnosis to a group of positive and negative variables the different open circuit faults of switching tube and winding can be distinguished by the diagnostic results. The experimental platform for fault diagnosis of six-phase permanent magnet fault-tolerant motor is built, and the experimental results verify the effectiveness and superiority of the proposed algorithm.