Samira Ben Salem

Support vector machine based decision for mechanical fault condition monitoring in induction motor using an advanced Hilbert-Park transform

In this work we suggest an original fault signature based on an improved combination of Hilbert and Park transforms. Starting from this combination we can create two fault signatures: Hilbert modulus current space vector (HMCSV) and Hilbert phase current space vector (HPCSV). These two fault signatures are subsequently analysed using the classical fast Fourier transform (FFT). The effects of mechanical faults on the HMCSV and HPCSV spectrums are described, and the related frequencies are determined. The magnitudes of spectral components, relative to the studied faults (air-gap eccentricity and outer raceway ball bearing defect), are extracted in order to develop the input vector necessary for learning and testing the support vector machine with an aim of classifying automatically the various states of the induction motor.

Smart wireless sensor networks for online faults diagnosis in induction machine

Display Omitted A new method has been proposed for online faults diagnosis in induction motors based on smart WSN combined with motor current signature analysis using FFT.The proposed method is novel as it is important to install low cost sensors and detection mechanisms along with induction machines to achieve short detection time and an automated way of reporting the fault.The system can distinguish a faulty motor from a healthy motor with a probability of 99% with less than 5% of false alarm.Simulation results presented show the efficiency of the proposed method to detect faults in induction machine. Online induction machine faults diagnosis is a concern to guarantee the overall production process efficiency. Nowadays, the industry demands the integration of smart wireless sensors networks (WSN) to improve the fault detection in order to reduce cost, maintenance and power consumption. Induction motors can develop one or more faults at the same time that can produce sever damages. The origin of most recurrent faults in rotary machines is in the components: stator, rotor, bearing and others. This work presents a novel methodology for the online faults diagnosis in induction motors. This technique uses the smart WSN to obtain the machine condition based on the motor stator current analysis. The implementation of the proposed smart sensor methodology allows the system to perform online fault detection in a fully automated way. Simulation results presented show the efficiency of the proposed method to detect simple and multiple faults in induction machine. It provides detailed analysis to address challenges in designing and deploying WSNs in industrial environments, and its reliability.

Induction motor mechanical fault identification using Park's vector approach

In this work we have shown that the extended Parks vector spectrum is rich in harmonics characteristics of mechanical defects (air-gap eccentricity and outer raceway bearing fault). About the use of Parks Lissajous curves to identify mechanical defects, we have demonstrated that this type of index can only detect the occurrence of a fault, but it cannot identify.

Induction motor fault diagnosis using an improved combination of Hilbert and Park transforms

In this work we propose an original failure signature based on an improved combination of Hilbert and Park transforms. Starting from this combination we can release two failure signatures: Hilbert modulus current space vector (HMCSV) and Hilbert phase current space vector (HPCSV). These two signatures are subsequently analysed using the classical fast Fourier transform (FFT). The effects of HMCSV and HPCSV spectrums are described and the related frequencies determined. This analysis offers an easy interpretation to conclude on the induction motor condition and its voltage supply state. The proposed signature shows its effectiveness and its robustness in both electrical and mechanical fault detection. This approach was applied to a 1.1 kw induction motor under normal operation and with the following fault: voltage unbalanced, broken rotor bar, air-gap eccentricity and ball bearing defect.

Induction motor fault diagnosis based on a Hilbert current space vector pattern analysis

In this work we propose an original failure signature based on the current Hilbert-Park vector pattern analysis. The advantage of this approach is which does not require a long temporal recording, and their processing is simple. This analysis offers an easy interpretation to conclude on the induction motor condition and its voltage supply state. The proposed signature shows its robustness and its power especially in the case of unloaded machine. This approach was applied to a 1.1 kw induction motor under normal operation and with the following fault: voltage unbalanced, broken rotor bar, air-gap eccentricity and ball bearing defect.

Support Vector Machine-Based Decision for Induction Motor Fault Diagnosis Using Air-Gap Torque Frequency

In this work we propose the air-gap torque as failure signature to detect mechanical faults in particular the eccentricity. In this way, we compare the proposed signature with those most used recently in particular the current space vector (Park vector) and complex apparent power. This signature is subsequently analysed using the classical fast Fourier transform (FFT). The magnitudes of spectral components relative to the studied fault are extracted in order to develop the input vector necessary for the pattern recognition tool based on support vector machine (SVM) approach with an aim of classifying automatically the various states of the induction motor. General Terms Diagnostic, Decision, Pattern recognition.

Experimental investigation of the eccentricity impact on the line current spectrum for induction motors fault diagnosis purposes

This paper deals with the use of the stator current signature analysis as a technique for the diagnostic of static, dynamic and mixed air-gap-eccentricity conditions in working three-phase induction motors. Associated spectra are analyzed and attribute fault harmonics are determined for an induction motor, under several types of air-gap eccentricity conditions. The effect of load levels in air-gap eccentricity fault detection is presented. The experimental results have revealed the potential of the spectral analysis of the stator current for the air-gap eccentricity fault detection.

Stray Flux analysis for monitoring eccentricity faults in induction motors: Experimental study

The machines air-gap cannot be perfectly smooth. Since a static or a dynamic eccentricity occurs, a mixed eccentricity behavior is observed and manifests by amplitude modulation of the stator current. Many other failures related to the motor condition or its driven load produce this same effect. Hence, by loading an induction machine, eccentricities incidence may be masked. Furthermore, because of power supply imperfections, the detection of characteristic frequencies around principal slot harmonics is not always practicable. In this paper, from theoretical development of eccentricities effects on analytical expressions of rotor and stator currents, a machine classification is proposed and related defect frequencies are predicted. The weakness of the motor current signature analysis is experimentally verified, and an alternative based on spectral analysis of the stray flux is proposed. Experimental results have shown an excellent capability of the suggested frequency signature for detecting and distinguishing the factual eccentricity.

Stator current signature analysis to monitor shaft misalignment in induction motor speed-controlled

In this work we discuss the efficiency of various stator current signatures for monitoring shaft misalignment that could affect mechanical systems when driven by an induction machine speed-controlled. In that respect, we have considered the Motor Current Signature Analysis, the Current Space Vector, and the Current Park Loci, under no-load as well as under full load conditions. On the other hand, the effect of the machines input-frequency on the detection efficiency is also discussed. Thus, we have assessed these three techniques when the motor is operating under full speeds condition and under half speeds condition. The presented results are experimentally validated on a laboratorys test-rig simulating an angular shaft misalignment.

Spur gearbox mixed fault detection using vibration envelope and motor stator current signatures analysis

Vibration analysis is the most used technique for gearbox fault diagnosis based on gear meshing frequency (GMF) magnitude tracking. The originality of this paper is illustrated in two steps. The meshing between two damaged teeth repeating to the Hunting tooth frequency (HTF) is clearly detected by performing spectral analysis to vibration envelope signal and the motor stator current, unlike the confirmation of [1][2] that this frequency is very low and cannot be measurable. In second step, gearbox input shaft break was predicted by the GMF harmonics magnitude comparison.