IEEE Transactions on Industrial Informatics · 2013 · 81 citations · 34 references
EngineeringElectrical EngineeringIndustrial EngineeringElectric MachinesPower ElectronicsTime Series AnalysisCondition MonitoringElectrical System DiagnosticsFault DetectionGenerator OperationInduction MachinesMagnetic Circuit AnalysisBroken Rotor Bar DetectionElectric Motor Fault DiagnosisEmpirical Mode DecompositionInduction Motor TheoryInduction MotorsMotor Fault DiagnosisOperating Point DeterminationRotor Bar Faults
This paper proposes a competent method for broken bars fault identification in line-start and inverter-fed industrial induction motors. The basis of the proposed method is to determine different frequency bands of the faulty motor current and extract a proper intrinsic mode function (IMF) developed by empirical mode decomposition (EMD). It is analytically proved that two specific IMFs, IMF <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> in line-start mode and IMF <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sub> in inverter-fed mode, can be used for precise prediction of the behavior of a faulty induction motor. The precision of the method is shown at different operating points, including different loads, fault, and reference speeds. The results indicate that the motor asymmetry is remarkably reduced at distributed broken bars. In addition, the proposed method is capable of determining the operating point of the faulty motor even at no/light loads. Finally, the Filippetti criterion is used, and a new aspect is analytically developed to justify the variations of IMF <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sub> with speed in an inverter-fed induction motor. The analytical results are verified by experiments.
34