Prony Method Estimation for Motor Current Signal Analysis Diagnostics in Rotor Cage Induction Motors

Abstract

This article presents an evaluation of Prony method and its implementation considerations for motor current signal analysis diagnostics in rotor cage induction motors. The broken rotor bar fault signature in current signals is evaluated using Prony method, where its advantages in comparison with fast Fourier transform are presented. The broken rotor bar fault signature could occur during the life cycle operation of induction motors, so that is why an effective early detection estimation technique of this fault could prevent an insulation failure or heavy damage, leaving the motor out of service. First, an overview of cage winding defects in rotor cage induction motors is presented. Next, Prony method and its considerations for the implementation in current signature analysis are described. Then, the performance of Prony method using numerical simulations is evaluated. Lastly, an assessment of Prony method as a tool for current signal analysis diagnostics is performed using a laboratory test system where real signals of an induction motor with broken rotor bar operated with/without a variable frequency drive are analyzed. The summary results of the estimation (amplitudes and frequencies) are presented in the results and discussion section.

Prony Method Estimation as a New Approach for Surge Comparison Testing in Turn Insulation Diagnostics for Three Phase Stator Windings

“…Recently, the Prony method has been used to obtain radar cross section (RCS) data which indicates the amount of signal reflected on an object by an electromagnetic wave applied by radar, one of the essential parameters for military aircraft design [32]. It has also been proposed and validated in motor current signal analysis application for broken rotor bar diagnostics [33]. The Prony method is proposed for this application, mainly because the surge waveforms obtained during an SCT are underdamped signals with minimum noise, as is observed in Figure 2, and the signal model used in Prony method estimation is a sum of exponentially decaying signals, as presented in (3).…”

Development of Broken Rotor Bar Fault Diagnosis Method with Sum of Weighted Fourier Series Coefficients Square

“…Energies 2022, 15, x FOR PEER REVIEW Based on the weighting factor in Figure 8, the experimental results of the sum of the weighted current coefficient ∑|γnCn| 2 are shown in Figure 9 for the time interval T of (a) 20, (b) 10, (c) 5, and (d) 4 s, where γ n = 2 f sl,rated / f s − n T from (10). The green line is the threshold level for diagnosing BRB faults.…”

“…However, many other decision rules can be implemented. ) from (10). The green line i the threshold level for diagnosing BRB faults.…”

“…However, the diagnosis of the BRB fault is challenging because there are very slight fault signals at the current, voltage, rotor speed, and vibration. To overcome this problem, several studies have investigated BRB fault diagnosis methods for induction motors [2][3][4][5][6][7][8][9][10][11][12][13][14][15][16][17]. The studied BRB fault diagnosis methods can be categorized into digital signal analysis, other information from special sensors, fault start cases, and neural networks or machine-learning-based algorithms [3].…”

“…In [5], Park transform in a synchronous reference frame was used to convert the fundamental AC component to DC. Otherwise, under time-varying conditions, short-period data-based methods, such as multiple signal classification (MUSIC) [6], discrete wavelet transform (DWT) [7,9], minimum description length (MDL) [8], and the Prony method [10] can be used. However, with a low slip frequency, they also require steady-state data with a long period to separate the specific fault signal components from the large fundamentals.…”