This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination. IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT 1 A Multiresolution Taylor–Kalman Approach for Broken Rotor Bar Detection in Cage Induction Motors Luis Alonso Trujillo-Guajardo, Johnny Rodriguez-Maldonado, M. A. Moonem, Member, IEEE, and Miguel Angel Platas-Garza , Member, IEEE Abstract— Broken rotor bar (BRB) fault is common in cage rotor induction motors. Motor current signature analy- sis (MCSA) has been a popular method to detect BRB faults. In the MCSA, the fault signature frequencies for BRB are close to the fundamental frequency. This spectral nearness leads to the requirement of larger observation windows, which inherently increases BRB fault detection time. This paper presents an alternative algorithm to detect BRB faults in induction motors from MCSA using two Taylor–Kalman (TK) filters in cascade with a subsampling scheme. The proposed BRB fault detection approach allows to use the TK filter to estimate lower frequencies with less computational burden in comparison to conventional TK analysis. Experimental analysis shows that nearly accurate estimates and competitive detection time can be achieved by the proposed BRB detection method. The performance of the proposed algorithm has been compared with classical spectral techniques using numerical simulations and records acquired from induction motors under real operating conditions. Index Terms—Broken rotor bar (BRB) fault signature, induc- tion motors, Kalman filter (KF), motor current signature analysis (MCSA), polynomial fit, Taylor–Kalman (TK) filter. I. I NTRODUCTION S INCE the early days of electric motors, induction motors have been placed as one of the most popular electrical machines, mainly for its ruggedness and low maintenance cost. Induction motors are subject to various faults, such as stator faults, broken rotor bar (BRB), bearing faults, mass unbalance, crawling, phase fault, and so on [1]. As per statistical studies sponsored by the IEEE and the Electric Power Research Institute, rotor faults are counted as 8%–9% of total induction motor faults [2]. BRB fault is very common in rotor cage induction motors (RCIMs). In an RCIM, rotor bars and end rings are typically made of alloys of either aluminum or copper [3]. When one or more rotor bars Manuscript received September 5, 2017; revised November 4, 2017; accepted November 20, 2017. The Associate Editor coordinating the review process was Dr. Datong Liu. (Corresponding author: Miguel Angel Platas-Garza.) L. A. Trujillo-Guajardo, J. Rodriguez-Maldonado, and M. A. Platas-Garza are with the Universidad Autónoma de Nuevo León, San Nicolas de Los Garza 66455, Mexico (e-mail: luis.trujillogjr@uanl.edu.mx; johnny. rodriguezml@uanl.edu.mx; miguel.platasgrz@uanl.edu.mx). M. A. Moonem is with Sandia National Laboratories, Albuquerque, NM 87123 USA (e-mail: m.a.moonem@gmail.com). Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/TIM.2018.2795895 get partially cracked or fully broken due to manufacturing defect, or excessive centrifugal stress at the slip ring or because of any other thermal or mechanical stress, the resulting fault is called as BRB fault. A broken bar or end ring does not carry current, which results in an unbalanced rotor flux that may cause unbalanced line currents, torque pulsation, and decreased average torque and excessive vibration [4]. If the fault is not detected and cleared in time, the former effects can cause an insulation failure or heavy damage, leaving the motor out of service. Some common physical indicators for BRBs in cage induc- tion motors are excessive vibration, noise, and sparking during motor starting, although there might be other defects that cause similar effects [5]. In early 1980s, a few fault detection techniques evolved [6]–[9]. Since then, research has been going on to detect and diagnose BRB faults in a noninvasive way [10]–[20]. Several methods and techniques have been presented in the literature to detect the accurate signature and diagnose the BRB [21]–[24]. Among these methods, motor current signature analysis (MCSA) has been popular to help diagnose problems, such as BRB, stator turn fault, shaft misalignment, eccentricity, and so on [25]–[30]. Many of these faults give rise to magnetic asymmetry in the rotor air gap, which produces spectral components at specific frequencies in the load current. The MCSA analysis is widely used in electric motor diagnostic online testers, as it does not require additional costly sensors rather than voltage and current probes as other noninvasive techniques like vibration analysis or ther- mography [31]–[35]. However, all these techniques could benefit predictive maintenance programs for RCIMs, espe- cially if they are used together. When MCSA is used, the tester is able to record voltage and current signals in high resolution. Usually, the length of the signal observation window must be established a priori. Once the acquisition is finished, the tester software performs spectral analysis of the signals. Fast Fourier transform (FFT) [36] has been a very well-known and widely used technique that has been used in a different diagnostic equipment to detect and analyze fault signatures [37]. It should be mentioned that the BRB fault signature could not be detected in some scenarios, for example, if there are transients and load changes or if the observation window used is too short. Usually, fault signature frequencies for BRB are closer to the fundamental frequency. FFT online testers use 0018-9456 © 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.