In this paper, a new approach is proposed for eccentricity fault detection in induction motors and estimation of the exact severities of the fault components. By using the Kalman filter estimator, the presented method can estimate degrees of the static, dynamic, and mixed eccentricity faults in the induction motors. The Kalman filter is a robust estimator having high capability in estimating the state variables of a dynamic system. This filter has many practical applications in industrial and nonindustrial systems and can implement continues‐/discrete‐time dynamic systems varying linearly or nonlinearly. In this paper, due to the nonlinear and continuous nature of the induction motor, the unscented Kalman Bucy filter (UKBF), which is a nonlinear continuous‐time filter, is employed. At first, the model of the eccentric induction motor is simulated. Then, by using this model, the measured quantities and the Kalman filter, the eccentricity fault is detected, and the exact severities of the fault components (static, dynamic, and mixed) are estimated for a real motor that has been artificially made to be eccentric. According to the attained results, it is demonstrated that the conducted approach has high capability in estimating the exact severities of the fault components, and it shows efficient performance in all eccentricity states.

中文翻译：

---

使用无味卡尔曼滤波器的感应电动机气隙偏心故障诊断和估计

本文提出了一种新方法，用于感应电动机的偏心故障检测和估计故障组件的严重程度。通过使用卡尔曼滤波器估计器，该方法可以估计感应电动机中静态，动态和混合偏心故障的程度。卡尔曼滤波器是一种鲁棒的估计器，在估计动态系统的状态变量方面具有很高的能力。该滤波器在工业和非工业系统中有许多实际应用，可以实现线性或非线性变化的连续/离散动态系统。在本文中，由于感应电动机的非线性和连续特性，因此采用了无味卡尔曼Bucy滤波器（UKBF），它是一个非线性连续时间滤波器。首先，模拟了偏心感应电动机的模型。然后，使用该模型，通过测量量和卡尔曼滤波器，检测出偏心故障，并为人工制造的实际电动机估算出故障分量的精确程度（静态，动态和混合）。偏心。根据获得的结果，证明了该方法在估计故障分量的精确度方面具有很高的能力，并且在所有偏心状态下都显示出有效的性能。