控制理论(社会学)
稳健性(进化)
卡尔曼滤波器
计算机科学
状态监测
多元微积分
故障检测与隔离
控制工程
扩展卡尔曼滤波器
断层(地质)
反电动势
观察员(物理)
容错
同步电动机
永磁同步电动机
伺服电动机
伺服
工程类
伺服驱动
伺服机构
可靠性(半导体)
国家观察员
规范化(社会学)
滤波器(信号处理)
作者
Yao Rao,Wei Wang,Wei Hua,Ming Cheng
标识
DOI:10.1109/tie.2026.3675226
摘要
Under extreme stress and dynamic servo operation, permanent magnet synchronous motor (PMSM) is more prone to interturn short-circuit fault (ISCF). Existing single fault detection methods provide only limited condition-monitoring information, and their steady state-based designs are also inadequate for capturing nonperiodic fault characteristics. To address this issue, this article proposes a back electromotive force-based multivariable condition monitoring (BEMF-MCM) method, which integrates fault detection, localization, and severity estimation under servo conditions. A dual-layer linear extended state observer (DL-LESO) is used to estimate thedq-axis back EMF and its fault components, which are further processed using negative-sequence transformation and dual normalization to extract stable fault characteristic. A variable-gain observer design improves robustness against motor state transients. Furthermore, an extended Kalman filter (EKF) model, incorporating attenuation coefficient and phase information, enables accurate fault severity estimation across all three phases under both healthy and faulty conditions. Extensive experimental results demonstrate the reliability and robustness of the proposed method.
科研通智能强力驱动
Strongly Powered by AbleSci AI