DSP implementation of a novel envelope analysis approach for the diagnosis of broken rotor bar in induction motor. (1st January 2014)

• Record Type: Journal Article
• Authors: Ahamed, Syed Kamruddin; Sarkar, Arghya; Mitra, Madhuchhanda; Sengupta, Samarjit
• Is Part Of: International journal of modelling, identification and control. Volume 22:Number 3(2014)
• Page Start: 275
• Page End: 284
• Publication Date: 2014-01-01
• Subjects: broken rotor bar -- digital signal processor -- DSP -- envelope -- Hilbert transform -- induction motor -- statistical parameters -- wavelet transform; Engineering -- Methodology -- Periodicals; Science -- Methodology -- Periodicals; 001.42
• Languages: English
• ISSNs: 1746-6172

Abstract

Design and implementation of a novel technique for the detection of broken rotor bar of the induction motor is presented. At first, the discrete wavelet transform (DWT) is performed on the windowed steady state stator currents and then discrete Hilbert transform is utilised to obtain envelopes of the detailed coefficients at higher wavelet level. It has been observed that the faulty motor produces higher statistical parameters – mean, rms, standard deviation and median of the envelopes than those for the healthy one. The method is implemented in real time by means of low cost TMS320VC5416 fixed point digital signal processor (DSP). A framework of assembly language program for the calculation of the DWT is given and the problems which should be paid attention during real time implementation are discussed. Illustrative laboratory test results confirm the validity and accurate performance of the proposed method.