Rough set based multi-class fault diagnosis of induction motor using Hilbert Transform

Konar, Pratyay; Bhawal, Shekhar; Saha, Moumita; Sil, Jaya; Chattopadhyay, Paramita

doi:10.1109/codis.2012.6422207

Cited by 4 publications

(2 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This gave us the ±π/2 phase-shift operator which is a basic property of the Hilbert transform [20]. A real function g(t) and its Hilbert transform ) ( t g are related to each other in such a way that they together create a so called strong analytic signal.…”

Section: Hilbert Transformmentioning

confidence: 99%

Mechanical fault diagnosis of induction motor using Hilbert pattern

Konar

Chattopadhyay

2013

2013 IEEE 1st International Conference on Condition Assessment Techniques in Electrical Systems (CATCON)

Self Cite

View full text Add to dashboard Cite

This paper deals with mechanical fault diagnosis in three-phase induction motor from radial vibration measurement. The Hilbert pattern of the 50 Hz mono-component signal extracted from the steady state vibration signature is analyzed and found to contain useful information needed for diagnosing different mechanicals faults. Since Hilbert transform can only be applied to a mono-component signal, Kaiser windowed FIR band pass filter is used to extract the monocomponent signal. Complex analytic signal is generated by using the mono-component signal as the real part and it's Hilbert Transform as the imaginary part. The concept of Hilbert transform for extraction of the instantaneous amplitude and frequency is utilized to extract important fault information from the non-stationary vibration signal and found to be quite efficient. This method does not require the analysis of fault frequency components which are slip dependent. Finally, an automatic diagnosis algorithm is attempted using SVM. The proposed method is almost independent of loading condition of the motor and has consistent performance even in presence of high level of noise.

show abstract

Section: Hilbert Transformmentioning

confidence: 99%