17 December 2020 in International Transactions on Electrical Energy Systems
The diagnosis of electrical motor for broken rotor bars faults through frequency signatures of the stator current method has been widely explored. In fact, a high‐resolution spectrum analysis based on the multiple signal classification (MUSIC) method has been widely applied to detect induction machine faults for that purpose. Moreover, MUSIC is a powerful tool that extracts significant frequencies from the signal, but cannot give precise amplitude information of the fault indicator frequency. Nevertheless, the MUSIC algorithm takes a long‐time acquisition to find many frequencies. To solve these problems, the short‐time least square Prony's technique is developed. The technique is based on a modified Prony's method for the estimation of amplitudes certain harmonics and frequencies of broken rotor bar fault. The main objective in this context is the reduction in the samples of measurement data with shorter time to allow a clearer view of the fault harmonics components when the motor turns at low load. It allows determining the characteristic frequencies and amplitudes of the lateral bands around the fundamental frequency component with great precision. Additionally, using this proposed method for the magnitude modulation of stator current envelope spectrum is produced and induces current harmonics by the rotor fault at frequency. It is based on tracking the amplitude of harmonics of the stator current envelope. The effectiveness of the proposed method is tested in simulation and a real‐time implementation on experimental validation.
20 December 2021 in International Journal of Modelling and Simulation
This paper presents a method based on advanced signal processing techniques for the detection of inte-rturn short circuit faults in an induction machine. The techniques presented can be classified into power spectral density estimation techniques and time-frequency techniques. The theoretical principles of these methods are presented and their performances are compared from the stator current analysis (MCSA). This article proposes a comparative study of these signal processing techniques. One is based on the power spectral density estimation techniques, such as Periodogram and Welch periodogram. The second method is based on the Hilbert transform (HT) to extract the envelope for the stator current. Then, this signal is processed via discrete wavelet transform (DWT) for determining the faulty components in the spectrum of the stator current envelope and identifying the eigenvalues of energy levels (HDWT). This technique is powerful for the non stationary signal processing. Moreover, the combination HT-DWT has assured its ability to evaluate the severity in detecting faults of the shorted turns and it is effective for nonstationary and stationary signals to extract useful information for the detection of ITSC fault. The techniques have usd experimentally data issued from the induction motor.
26 February 2021 in International Journal of Dynamics and Control
19 December 2022 in International Journal of Emerging Electric Power Systems
This paper presents the development of a new algorithm for the diagnosis of induction motor inter-turn short circuit and mixed combined fault detection based on the reduction Park’s vector modulus and discrete wavelet transform. The proposed method is based on a combination between the RPVM approach and DWT. This new signal of the RPVM is a cost-effective alternative implementation of the Extended Park’s Vector approach through monitoring only the fluctuations of the stator current zero-crossing times in the frequency domain. Using this technique, the characteristic frequency components of the fault are extracted from only one phase stator current when compared to Park’s vector modulus and Hilbert modulus. The main advantages of the RPVM signal are to acquire the values of a single-phase stator current to allow reducing the calculation complexity and for extracting features frequencies of the faults. Thus, the reduced amount of data samples must be stored and processed. This study investigates theoretically by using the proposed RPVM signal for performing to diagnose induction motor faults through the DWT method and processed via energy eigenvalue of energies. This proposed approach RPVM-DWT is validated experimentally to demonstrate its effectiveness in that the proposed algorithm reduces the computational cost and can be used in induction motor fault detections and identifications with different fault severity conditions.