Motor current signature analysis of incipient broken rotor bar of squirrel cage induction motor

Nowadays, manufacturing companies are making great efforts to develop incipient fault detection, as it prevents the unscheduled downtime and hence reduces maintenance costs. The machine problem and irregularity can be detected at an early stage using a suitable condition monitoring. The condition monitoring schemes have concentrated on sensing specific failure modes in different parts of the motor. Rotor faults are of significance importance as they cause secondary failures which lead to serious motor malfunctions. Detection of rotor faults has long been an important but difficult job in the detection area of motor faults. Motor current signature analysis (MCSA) is considered as an effective condition monitoring in any induction motor. However, a signal processing technique, which enhances the fault signature and suppress the dominant system dynamics and noise must be considered. Previous researches found that when broken bars occur in the machine rotor, the anomaly of electromagnetic field in the air gap will cause two sideband frequency components presented in the stator current spectrum. Therefore, identification of these sideband frequencies can be used as a convenient and reliable approach to detect the broken rotor bar in induction machines. Frequency analysis as well as time-frequency analysis is the most common signal processing methods applied for fault detection of induction motor. In this research, the effectiveness of these two analysis techniques were investigated for incipient broken rotor bar detection in squirrel-cage induction motor under different levels of load. The result showed that frequency analysis of current signal cannot provide accurate information for incipient fault detection. Therefore, time-frequency analysis was examined for incipient broken rotor bar detection. Wavelet transform of the raw signal depends on the type of wavelet function used for decomposition is different. In view of that, it is desirable to select the appropriate wavelet function, which produces the best results for the signal being analyzed according to the purpose of the research. Therefore, this research investigated the analysis of current signal using different wavelet functions for effective and incipient detection of broken rotor bar in squirrel-cage induction motor. Different functions, namely, Biorthogonal, Coiflet, Daubechies, were compared in screening the features corresponding to the fault present in motor. Among those wavelet functions studied, Daubechies1 provided much more reliable information for incipient detection of broken rotor bar.

Preview PDF FK 2011 62R.pdf Download (853kB) | Preview

Actions (login required)

View Item