Abstract

Fault detection in an induction motor, particularly at premature stage has become necessary to avoid unexpected damage in industrial process. In this paper, an approach to detect the early stage faults in induction machine using motor current signature analysis (MCSA) is presented. It is proposed to estimate the fault severity from stator current using noise cancelation by an adaptive filter (Wiener filter). Wavelet De-noising technique is implemented to reduce the effect of noise floor in noise canceled stator current. Different categories of bearing faults, broken rotor fault and stator inter turn faults in induction motor are estimated with and without de-nosing using pre-fault component cancelation (Noise cancelation). In addition, fault index based on standard deviation (SD) and simple square integral (SSI) value of noise canceled stator current are proposed. The proposed fault detection topology is examined using simulations and experiments on a 3HP, 1HP and 0.5HP induction motors for bearing, broken rotor and stator inter turn faults respectively.

摘要

感应电机中的故障检测，尤其是在过早阶段，已成为避免工业过程中意外损坏的必要条件。在本文中，提出了一种使用电机电流特征分析 (MCSA) 检测感应电机早期故障的方法。建议通过自适应滤波器（维纳滤波器）使用噪声消除来估计定子电流的故障严重程度。实施小波去噪技术以减少噪声消除定子电流中噪声本底的影响。感应电机中不同类别的轴承故障、转子断裂故障和定子匝间故障可以使用故障前分量消除（噪声消除）进行去噪和不去噪的估计。此外，提出了基于噪声消除定子电流标准差(SD)和简单平方积分(SSI)值的故障指标。所提出的故障检测拓扑分别使用 3HP、1HP 和 0.5HP 感应电机的模拟和实验来检查轴承、转子损坏和定子匝间故障。