Shaeboub, Abdulkarim (2018) The Monitoring of Induction Machines Using Electrical Signals from the Variable Speed Drive. Doctoral thesis, University of Huddersfield.

Induction motors are the most widely used industrial prime movers, mainly because of their simple yet powerful construction, ergonomic adaptability, rugged and highly robust structure combined with high reliability. However, under extreme and complex operations, such motors are subject to premature faults, which can be more significant when variable speed drive (VSDs) are used, due to the presence of more voltage harmonics, spikes and increases in operating temperature. In addition, VSD based systems also cause more noise in measured instantaneous current signals. These make it more difficult to investigate and accurately diagnose system faults in order to keep VSD based motors operating at an optimal level and avoid excessive energy consumption and damage to system.

However, insufficient work has been carried out exploring fault diagnosis using terminal voltage and motor current signals of VSD motors which are increasingly used in industry. To fill these gaps, this thesis investigates the detection of stator and rotor faults (i.e. shorted turn faults, open-circuit faults, broken rotor bars, and stator winding asymmetry combined with broken rotor bar faults) using electrical signals from VSDs under different loads and different speeds conditions.

Evaluation results show that under open loop control mode, both stator and rotor faults cause an increase in the amplitude of sidebands of the motor current signature. However, no changes were found that could be used for fault detection in the motor voltage signature with respect to open loop control mode. This is because, when the drive is in open-loop operation, there is no feedback to the drive and torque oscillations modulate the motor current only. The V/Hz ratio is kept constant even when the slip changes either due to the load or the fault.

On the other hand, the increase in the sideband amplitude can be observed in both the current and voltage signals under the sensorless control mode with the voltage spectrum demonstrating a slightly better performance than the motor current spectrum, because the VSD regulates the voltage to adapt changes in the electromagnetic torque caused by the faults. The comparative results between current and voltage spectra under both control modes show that the sensorless control gives more reliable diagnosis.

In order to monitor the condition of electrical drives accuratly and effectively, demodulation analysis such as modulation signal bispectrum (MSB) of the electrical signals from the VSDs has been explored extensively in this thesis to detect and diagnose different motor faults. MSB analysis has been shown to provide good noise reduction, and more accurate and reliable diagnosis. It gives a more correct indication of the fault severity and fault location for all operating conditions.

This study also examines detecting and diagnosing the effect of an asymmetric stator winding combined with broken rotor bar (BRB) faults under the sensorless operation mode. It examines the effectiveness of conventional diagnostic features in both motor current and voltage signals using power spectrum (PS) and MSB analysis. The obtained results show that the combined fault causes an additional increase in the sideband amplitude and this increase can be observed in both the current and voltage signals.

The MSB diagnosis based on the voltage signals is more sensitive to detect motor faults at lower loads compared with that of current signals. Moreover, this research presented a new method based on MSB sideband estimation (MSB-SE). It is shown that using MSB-SE, the sidebands due to weak fault signatures can be quantified more accurately, which results in more consistent detection and accurate diagnosis of the fault severity.

Abdulkarim Shaeboub FINAL THESIS.PDF - Accepted Version
Available under License Creative Commons Attribution Non-commercial No Derivatives.

Download (4MB) | Preview


Downloads per month over past year

Add to AnyAdd to TwitterAdd to FacebookAdd to LinkedinAdd to PinterestAdd to Email