This chapter presents an overview of faults found in electrical machines and their diagnosis, with a special reference to induction machines and their fault detection. It briefly outlines the composition of the induction machine. This description enables us to better understand induction machine failures in their physical dimensions. From a mechanical point of view, induction machines can be composed of three distinct parts: the stator, the rotor and the bearings. Although the induction machine is said to be robust, it may sometimes present different types of faults. These faults are found in the different parts of the machine, starting with the stator phase connection and finishing with the mechanical coupling between the rotating shaft and the load. These failures can be predicted or unexpected, mechanical, electrical or magnetic, and they have very different causes. The chapter presents an overview of diagnosis methods applied to electrical machines. electric machines; fault diagnosis

Request Full-text Paper PDF

To read the full-text of this research,
you can request a copy directly from the author.

... To this end, two kinds of methods have been developed: the model-based [13] and the data-driven approaches [14]. Each of them has its own advantages and drawbacks, and many have already been tested for the online diagnosis of variable speed drives [15] [17], batteries and fuel cells [18] [20]. For supercapacitors, attention has so far been focused on aging characterization and diagnosis by parameter estimation [21] [25], but rarely on taking the context (constraints) of their in situ diagnosis into account. ...

... 1) Continuous-time state-space representation: The discrete recurrence equations and identification algorithms derived from the least squares framework are often used for the synthesis of control laws [27]. But if the goal is to deeply understand the system behavior for its diagnosis and prognosis, then the identification of knowledge continuous-time models presents many attractive features, in particularly because their parameters have a physical meaning and can be more simply linked to the physical phenomena occurring in the system being monitored [15]. Then, the physical ordinary differential equations of the system naturally lead to a continuous-time state-space representation: ...

... A resistance R p paralleled to the capacitance C o usually models its energy losses. But the typical value of R p is several M Ω, and identifying models whose parameters have very different orders of magnitude may lead to many numerical problems [15]. This will be discussed in section V. ...

Electrical double-layer capacitors (EDLCs) are now widely used as power buffers in energy storage systems, especially for electrical transport applications. Because of their high acquisition costs, their predictive maintenance represents a key issue for the development of electrical traction systems. This paper investigates the diagnosis of EDLCs by a parametric estimation of a dynamic continuous-time model. Indeed, the physical interpretation of their parameters simplifies the supercapacitor diagnosis and prognosis. Due to the complex electrochemical phenomena occurring in EDLC cells, modeling errors are far more important than measurement noise. Therefore, we propose to use the output error minimization techniques and the relative sensitivity functions to evaluate and compare the diagnosis potential of several model structures. This allows the use of some comprehensive comparison criteria without any assumption about the statistical characteristics of the system disturbances. An experimental accelerated aging process of 200 commercial super capacitors (Nichicon, 1F, 20 weeks long aging at 65 ◦C, 0 and 2.5 V dc polarization) is also performed to make statistics and illustrate the methodology.

... The processes of starting an electromechanical system (EMS) are accompanied by peak power consumption, signif- icant influence on the service life of electromechanical and processing equipment. Starting powerful electric drives, espe- cially in commensurable capacity power supply networks, has a significant impact on a mains supply [2]. Starting voltage 7. Zvit NDR za temoyu gospdogovoru z TOV NVP "Adioz" (m. ...

... One of the most promising approaches to the solution of the problem of estimating the starting process parameters and comparing design solutions of starting systems is the use of the starting process efficiency index [2], based on the re- source conversion efficiency theory [3]. However, the known solutions do not take into account the continuous nature of resource consumption during the starting process, consider- ing it as a simple lumped-parameter operation. ...

  • Tytiuk Valery Tytiuk Valery

Currently, an approach based on the lumped-parameter resource consumption model is used to determine the starting process efficiency index. This is due to the simplicity of analytical expressions for the efficiency index. In actual practice, the resource consumption of the starting process is time-phased in nature, which may lead to biased estimates of the efficiency index and inaccurate operation of the optimization system. In the paper, the actual form of the signal of the cost estimate of input products over time was determined using mathematical modeling of the controlled start system. It is proposed to use a fractional rational function to approximate these signals. The parameters of this approximation at different values of the control action were obtained using the Matlab Curve Fitting Toolbox. The analytical expressions for determining the resource consumption, potential effect and efficiency index of the starting process were obtained on the basis of the proposed approximation. It was found that the position of the maximum efficiency index on the x-axis, obtained when using the lumped-parameter and distributed-parameter starting operation models varies slightly (within the 5 % margin of error). This will reasonably facilitate hardware implementation of optimal control systems of starting processes through the application of the lumped-parameter starting operation model.

... Fig. 3 shows the technical so for the in-situ excitation of the insulation frequency excitation is applied to the ge through a coupling box, between one or seve stator housing. Therefore, the system performed in an open-loop context, and d specific algorithms [17]. ...

... r Method used for system ature of the model arameters and/or to nd structure noises, gnose, or even the acceptable computational cost [ In fact, the insulation sys several years and the calcul monitoring system is clearly n economic and industrial st Moreover, the outputs of contin R-L-C networks are not linear Thus, we propose to use the sensitivity functions for the i [21]. This method presents a c than methods derived from le unbiased estimator with a re modeling errors [17]. This is an required simplicity of the diagn to an important modeling noise Fig 5 resumes ...

This paper investigates the online monitoring of electrical machine winding insulation systems based on the parametric modeling and identification. The proposed method consists in monitoring the drift of diagnostic indicators built from in-situ estimation of high-frequency electrical model parameters. The involved model structures are derived from the RLC network modeling of the winding insulation. Because they often present an important modeling noise, we propose to use the output error method not only to estimate the model parameter values but also to evaluate their uncertainty. This approach is based on the numerical integration of the model sensitivity functions. The so-called global identification scheme is coupled with an optimization algorithm that brings the best combination of any diagnostic model structure and its excitation protocol usable in operating conditions. Experimental data recorded from an industrial wound machines are used to illustrate the methodology.

... Additionally, the rapidly evolving industries and the increasing demand for hybrid and electric vehicles indicate that there will be a further increase in this rate of usage. Electrical or mechanical faults will occur during lifetime of an electrical machine [2] and may lead to catastrophic failures. Fault detection and diagnosis is a crucial task to prevent and predict these undesirable failures that can lead to unscheduled and costly downtime. ...

The most frequent faults in rotating electrical machines occur in their rolling element bearings. Thus, an effective health diagnosis mechanism of rolling element bearings is necessary from operational and economical points of view. Recently, convolutional neural networks (CNNs) have been proposed for bearing fault detection and identification. However, two major drawbacks of these models are (a) their lack of ability to capture global information about the input vector and to derive knowledge about the statistical properties of the latter and (b) the high demand for computational resources. In this paper, short time Fourier transform (STFT) is proposed as a pre-processing step to acquire time-frequency representation vibration images from raw data in variable healthy or faulty conditions. To diagnose and classify the vibration images, the image classification transformer (ICT), inspired from the transformers used for natural language processing, has been suitably adapted to work as an image classifier trained in a supervised manner and is also proposed as an alternative method to CNNs. Simulation results on a famous and well-established rolling element bearing fault detection benchmark show the effectiveness of the proposed method, which achieved 98.3% accuracy (on the test dataset) while requiring substantially fewer computational resources to be trained compared to the CNN approach.

... Widespread growth of use of electric machines in industries has attracted many researchers to work in this arena. Author [1] has given an overview of faults usually found in electrical machines and their diagnosis with special emphasis on Induction Machines. Authors [2] have elaborated about different faults and their causes; then various instruments required for acquisition of different signals required in condition analysis of machines. ...

Condition monitoring of induction motors has attracted the researchers' interest in recent years with increasing significance of rotary machines in the industries. Among all induction motor faults, bearing needs the utmost attention being the most critical component of any machine. Several researchers have implemented numerous techniques in efforts to predictive maintenance of bearings. For the diagnosis and detection of both localized and distributed effects, many researchers have given the priority to vibration signals as the measure of bearing health status. Many significant works have been used vibration signatures available from Case Western Reserve University bearing data center. This paper attempts to survey and summarize the recent research developments in the fault diagnosis of bearings, which can be useful for researchers to have a look at latest advancements in this field. Various signal processing techniques, feature extraction and selection techniques and classification techniques used in the concerned research area have been briefed in this paper. Various eminent works of the concerned field have been outlined here along with their comprehensive analysis and the possible enhancements in the diagnostic procedures.

... (Ricardo Alvarez-Salas) leyes de control y dispositivos semiconductores más versátiles (Mujica and Espinoza-Pérez (2014)). Sin embargo, la mayoría de los algoritmos de control pueden no ser confiables cuando ocurren fallas (Blanke et al. (2003)), (Puig et al. (2004)), (Trigeassou (2011)), (Toliyat et al. (2013)), (Verde et al. (2013)). Las fallas en el circuito de estator del MI trifásico representan un porcentaje significativo de las fallas en el motor (de Angelo et al. (2007)). ...

  • Ricardo Alvarez-Salas Ricardo Alvarez-Salas

En este trabajo se presenta un conjunto de herramientas basadas en la transformada rapida de Fourier (TRF) y la transformada discreta ondeleta (TDO) Haar, Daubechies y Coiflet utilizando el modulo del vector espacial instantaneo (VEI) de las corrientes de estator del motor de induccion trifasico para diagnosticar fallas en el estator y en el rotor. Se efectua un estudio en simulacion y una validacion experimental de fallas electricas del estator para cortocircuito entre espiras de un mismo devanado y problemas de fisura de barra y anillo de cortocircuito de rotor. El objetivo del trabajo es contar con una herramienta de tipo no invasiva que permita efectuar el diagnostico de fallas incipientes (con un caracter preventivo) con el fin de evitar perdidas economicas importantes en diferentes procesos industriales que utilizan maquinas de induccion aprovechando para ello las virtudes que ofrece la TRF y la TDO mediante el analisis de las firmas de corriente del estator del motor de induccion (MI), asi como el diagnostico de fallas drasticas con fines correctivos.

... Moreover, both internal and external faults refer to components of electrical and/or mechanical part of the motor. It is worth to emphasize that it is investigated that more than 40% of induction machine failures concern mechanical subtype (Trigeassou, 2011). They are mainly bearing, rotor and stator failures (Akin & Rahimian, 2013; Choi, 2013; Thorsen & Dalva, 1995; Bonnett & Yung, 2008). ...

  • Beata Mrugalska Beata Mrugalska

In: Constantine Stephanidis & Margherita Antona (eds.), Universal Access in Human-Computer Interaction. Design for All and Accessibility Practice. 8th International Conference, UAHCI 2014, Held as Part of HCI International 2014, Heraklion, Crete, Greece, 22-27 June 2014, Proceedings, Part IV, LNCS 8516: 237-245. In order to obtain high performance and safety degree in machine operation it is necessary to identify potential machine faults. Such faults may have an effect on the machine itself but may also lead to accidents at work. Thus, the activities aiming at the identification of the relation between the failure causes and their effects seem to be necessary to be undertaken. With this purpose, the first part of the paper concerns the analysis of machine faults. In order to achieve it, induction machines are chosen and are widely discussed. In the next part of it the issues of occupational risk are presented, in particularly statistical data about accidents at work in Polish enterprises. It allows to determine the problem of faults and their influence on workers' safety in industrial environment.

The three-phase induction motor is the main machine used for electromechanical energy conversion, due to its consolidated construction characteristics. As a consequence of its great importance and industrial application, researches in the fault identification area are constantly conducted to reduce the maintenance rate and the losses, during the productive process, caused by undesirable downtime. In this sense, this work proposes an alternative methodology, based on the differential evolution algorithm, to identify stator short-circuit fault in induction motors connected directly to the electrical grid, using voltage and current signals in time domain. The differential evolution algorithm is used to estimate the electrical parameters of the induction motor, based on the model of the equivalent electrical circuit. Stator fault is identified by calculating the variation of the estimated magnetizing inductance of the motor under no fault condition. The proposed method is validated through experimental tests on 1 HP and 2 HP motors under conditions of load torque variation and unbalanced voltages.

This paper presents a simple state-space representation of induction machine (IM) transients under stator winding inter-turn short-circuit fault. The mathematical development of this dynamic model is presented in detail and the differential equations are transformed into space-vector form for reduction reasons. The simulations of this transient model are conducted in MatLab script using ordinary differential equation (ODE) solver. For verification purposes, the results of this simulation are compared with the ones obtained for the symmetrical IM. A turn fault signature study is performed for fault detection and identification using the electromagnetic torque ripple and the symmetric components transformation of the stator currents.

  • Валерий Константинович Тытюк

The operation of most manufacturing processes begins with the electric drive startup, which determines the importance and urgency of optimization of EMS startup processes. In determining the objective function of optimum systems of controlled EMS startup, it was proposed to use the principle of focus on the value added. To implement this approach, various types of input and output products of the EMS startup process were investigated. It is shown that in the analysis of startup processes, it is important to consider accelerated wear of the electromechanical equipment, which is an essential component of the input products of the startup process. The design formulas and models for determining the estimates of input and output products of the EMS startup process, the startup duration were proposed. To overcome the dimensionality problem when dealing with heterogeneous products, it is advisable to use cost estimated of these products. This approach is an economically sound kind of the expert evaluation method. Based on the known efficiency indicator of resource conversion, an expression of the startup efficiency indicator, taking into account the potential economic risks of the startup failure was obtained. By the mathematical modeling methods using Matlab/Simulink, dependences of power consumption, wear and efficiency indicators of the EMS startup process on the control action were obtained. The extreme nature of the dependence of the startup process efficiency indicator on the control action was shown. The results can be used in the synthesis of the optimal control systems of the EMS startup process.

  • Aleksandrs Gasparjans
  • Aleksandrs Terebkovs
  • Anastasia Zhiravetska

A method of technical diagnostics of ship diesel engine – generator installation – is proposed. Spectral-power diagnostic parameters of the synchronous generator voltage and currents are used. The electric machine in this case is the multipurpose sensor of diagnostic parameters. A judgment on the quality of the operational processes in diesel engine cylinders and its technical condition is possible on the basis of these parameters. This method is applicable to piston compressor installations with electric drive. On the basis of such parameters as rotating torque, angular speed and angular acceleration it is possible to estimate the quality of the operating process in the cylinders of a diesel engine, the condition of its cylinder-piston group and the crank gear mechanism. The investigation was realized on the basis of a diesel-generator with linear load. The generator operation was considered for the case of constant RL load. Together with the above mentioned, the condition of bearings of synchronous machines, uniformity of the air gap, windings of the electric machine were estimated during the experiments as well. The frequency spectrum of the stator current of the generator was researched and analyzed. In this case the synchronous machine is becoming a rather exact multipurpose diagnostic sensor. The signal of non-uniformity in the operation process of diesel engine cylinders and its technical condition is the increasing of the amplitudes of typical frequencies.

Change detection problem refers to the detection of abnormal regimes and incipient faults of electrical circuits, devices or equipments. The problem is coming more and more important as the requirements for high and constant quality, low maintenance and high reliability of complex electronic products are continuously increasing. Classical methods use statistical signal processing techniques, various frequency transforms and techniques based on models. In all these approaches the moment of change detection must be known as precisely as possible. This work presents preliminary results in solving change detection problem based on two methods. The first one is based on signal modeling and detects changes by observing the changes in the model's parameters. The second one is based on Renyi entropy processing and it uses direct, simple and efficient parameter estimation methods to describe the circuit's behavior. Both methods could serve as references for other more advanced techniques, e.g. based on frequency transforms and complex optimization criteria.

  • Estelle Cherrier Estelle Cherrier

De façon générale, cette thèse porte sur l'estimation de l'état et des entrées inconnues pour une classe de systèmes non linéaires. De façon plus particulière, le problème est abordé sous l'angle de la conception d'un système de transmission sécurisée d'informations exploitant les propriétés des systèmes chaotiques et leur capacité de synchronisation. Les travaux présentés traitent trois points principaux, à savoir le choix de l'émetteur, le développement du récepteur, et la mise au point du processus de transmission de l'information ou du message. L'émetteur retenu est un système non linéaire chaotique dont la dynamique comporte un retard, ce qui lui confère un comportement particulièrement complexe. La conception du récepteur repose sur la synthèse d'un observateur non linéaire, dont la stabilité et la convergence garantissent la synchronisation avec l'émetteur. L'insertion du message est réalisée par modulation de la phase d'un signal porteur chaotique. Le décryptage de l'information s'apparente à une restauration d'entrée inconnue au niveau du récepteur. Une étude de la sécurité du processus de cryptage/décryptage a été menée, reposant sur des techniques standard de cryptanalyse. Des multimodèles chaotiques ont été proposés pour renforcer la sécurité du processus de synchronisation

Condition monitoring leading to fault diagnosis and prediction of electrical machines and drives has recently become of importance. The topic has attracted researchers to work in during the past few years because of its great influence on the operational continuation of many industrial processes. Correct diagnosis and early detection of incipient faults result in fast unscheduled maintenance and short down time for the machine under consideration. It also avoids harmful, sometimes devastative, consequences and helps reduce financial loss. Reduction of the human experts involvement in the diagnosis process has gradually taken place upon the recent developments in the modern artificial intelligence (AI) tools. Artificial neural networks (ANNs), fuzzy and adaptive fuzzy systems, and expert systems are good candidates for the automation of the diagnostic procedures. This present work surveys the principles and criteria of the diagnosis process. It introduces the current research achievements to apply AI techniques in the diagnostic systems of electrical machines and drives.

This paper investigates the use of fuzzy logic for fault detection and diagnosis in a pulsewidth modulation voltage source inverter (PWM-VSI) induction motor drive. The proposed fuzzy technique requires the measurement of the output inverter currents to detect intermittent loss of firing pulses in the inverter power switches. For diagnosis purposes, a localization domain made with seven patterns is built with the stator Concordia current vector. One is dedicated to the healthy domain and the six others to each inverter power switch. The fuzzy bases of the proposed technique are extracted from the current analysis of the fault modes in the PWM-VSI. Experimental results on a 1.5-kW induction motor drive are presented to demonstrate the effectiveness of the proposed fuzzy approach.

Motor fault detection and diagnosis involves processing a large amount of information of the motor system. With the combined synergy of fuzzy logic and neural networks, a better understanding of the heuristics underlying the motor fault detection/diagnosis process and successful fault detection/diagnosis schemes can be achieved. This paper presents two neural fuzzy (NN/FZ) inference systems, namely, fuzzy adaptive learning control/decision network (FALCON) and adaptive network based fuzzy inference system (ANFIS), with applications to induction motor fault detection/diagnosis problems. The general specifications of the NN/FZ systems are discussed. In addition, the fault detection/diagnosis structures are analyzed and compared with regard to their learning algorithms, initial knowledge requirements, extracted knowledge types, domain partitioning, rule structuring and modifications. Simulated experimental results are presented in terms of motor fault detection accuracy and knowledge extraction feasibility. Results suggest new and promising research areas for using NN/FZ inference systems for incipient fault detection and diagnosis in induction motors

  • David Luenberger David Luenberger

Observers which approximately reconstruct missing state-variable information necessary for control are presented in an introductory manner. The special topics of the identity observer, a reduced-order observer, linear functional observers, stability properties, and dual observers are discussed.

An observer for nonlinear systems is constructed under rather general technical assumptions (the fact that some functions are globally Lipschitz). This observer works either for autonomous systems or for nonlinear systems that are observable for any input. A tentative application to biological systems is described

  • G. Comtet-Varga
  • C. Christophe
  • Vincent Cocquempot Vincent Cocquempot
  • M. Staroswiecki

This paper is concerned with Fault Detection and Isolation for induction motors. A new procedure for the calculation of the ARR (Analytical Redundancy Relations) is presented. This method is based on the elimination of the unknown variables (states, unknown inputs,...) in the model equations. Simulation results are presented in normal situation and in presence of a speed sensor fault.

  • Sekhar Raghavan
  • J. Karl Hedrick

A viable design methodology to construct observers for a class of nonlinear systems is developed. The proposed technique is based on the off-line solution of a Riccati equation, and can be solved using commercially available software packages. For globally valid results, the class of systems considered is characterized by globally Lipschitz nonlinearities. Local results relax this assumption to only a local requirement. For a more general description of nonlinear systems, the methodology yields approximate observers, locally. The proposed theory is used to design an observer for a single-link flexible joint robot. This observer estimates the robot link variables based on the joint measurements.

  • Saad Al kazzaz Saad Al kazzaz
  • G.K. Singh

Condition monitoring is used for increasing machinery availability and machinery performance, reducing consequential damage, increasing machine life, reducing spare parts inventories, and reducing breakdown maintenance. An efficient condition monitoring scheme is capable of providing warning and predicting the faults at early stages. The monitoring system obtains information about the machine in the form of primary data and through the use of modern signal processing techniques; it is possible to give vital information to equipment operator before it catastrophically fails. The suitability of a signal processing technique to be used depends upon the nature of the signal and the required accuracy of the obtained information. Therefore, in this paper, signals obtained from the monitoring system are treated with different processing techniques with suitably modified algorithms to extract detailed information for machine health diagnosis. In this study, on-line analysis of the acquired signals has been performed using c++, while matlab has been used to perform the off-line analysis.

  • Rudolf E Kalman Rudolf E Kalman

The classical filtering and prediction problem is re-examined using the Bode-Sliannon representation of random processes and the "state-transition" method of analysis of dynamic systems. New results are: (1) The formulation and methods of solution of the problem apply without modification to stationary and nonstationary statistics and to growing-memory and infinitememory filters. (2) A nonlinear difference (or differential) equation is derived for the covariance matrix of the optimal estimation error. From the solution of this equation the coefficients of the difference (or differential) equation of the optimal linear filter are obtained without further calculations. (3) The filtering problem is shown to be the dual of the noise-free regulator problem. The new method developed here is applied to two well-known problems, confirming and extending earlier results. The discussion is largely self-contained and proceeds from first principles; basic concepts of the theory of random processes are reviewed in the Appendix.

  • Peter C. Young Peter C. Young

The paper reviews the progress of research on parameter estimation for continuous-time models of dynamic systems over the period 1958–1980. Major developments are considered in historical order and within a classification system which conforms as closely as possible to that which has arisen naturally over the past two decades. While every attempt is made to consider progress in other related scientific disciplines, such as econometrics, the major accent is on research in the control and systems field.

  • Rolf Isermann

For further improvement of the reliability and safety of machines the automatic early detection and localization of faults is of high interest. The conventional approach is to monitor some important variables like temperature, pressure, vibration and to generate alarms if certain limits are exceeded. However, developing internal faults are in this way only detected at a rather late stage. By applying static and dynamic process models and common process input and output measurements the inherent relationships and redundancies can be used to detect faults earlier and to localize them better.Changes in process and signal parameters are very well suited for fault detection. The paper describes a general methodology for machines and other processes by using few measurements, dynamic process and signal models and parameter estimation to generate analytical symptoms. Heuristic symptoms observed by the operator or based on, e.g. the process history are another source for fault diagnosis. Both kinds of symptoms are weighted by confidence measures and are fed into a knowledge based fault diagnosis procedure. The diagnosis is performed by forward and backward chaining according to causal fault-symptom trees.The considered machines consist of a motor, a drive chain and a working process or load. They may be electrical motor or combustion engine driven pumps, fans or machine tools with gear or belt drive chains. As one example, experimental results are shown for a DC motor powered feed drive of a machine tool. A summary of practical results with other machines is also given.

  • Lennart Ljung Lennart Ljung

The sections in this article are1The Problem2Background and Literature3Outline4Displaying the Basic Ideas: Arx Models and the Linear Least Squares Method5Model Structures I: Linear Models6Model Structures Ii: Nonlinear Black-Box Models7General Parameter Estimation Techniques8Special Estimation Techniques for Linear Black-Box Models9Data Quality10Model Validation and Model Selection11Back to Data: The Practical Side of Identification

Most of faults in three-phase induction motors have relationship with airgap eccentricity. There are two forms of airgap eccentricity: static (SE) and dynamic (DE). According to the literatures, the well known signatures of dynamic eccentricity, on the stator current spectra, are sidebands around the principal slot harmonics (PSH). However, many other researches have shown that DE induces also spectral components around the fundamental, but few are reported on the sources and the causes of these components. In this direction and since it is difficult to study experimentally the DE separately from the SE; the present paper attempts to explain, analytically and by simulation, the generation process of all frequency components that are a function of only DE. For that reason, a detailed analytical study for three-phase induction motors working under DE is performed. This study is based on rotating field approach. A general theoretical analysis of the interaction between all harmonics of the eccentric airgap permeance and the stator and rotor MMF components is put forward. The simulation results, obtained from an accurate model, confirm the existence of specific frequency components around the fundamental, caused by the dynamic airgap eccentricity. The interactions between the DE and the inherent SE are also illustrated using this mathematical model.

  • O.V. Thorsen
  • M. Dalva

The study gives a synopsis over condition monitoring methods both as a diagnostic tool and as a technique for failure identification in high voltage induction motors in industry. New running experience data for 483 motor units with 6135 unit years are registered and processed statistically, to reveal the connection between motor data, protection and condition monitoring methods, maintenance philosophy and different types of failures. The different types of failures are further analyzed to failure-initiators, -contributors and -underlying causes. The results have been compared with those of a previous survey, IEEE Report of Large Motor Reliability Survey of Industrial and Commercial Installations, 1985. In the present survey the motors are in the range of 100 to 1300 kW, 47% of them between 100 and 500 kW.

  • Austin Bonnett
  • Chuck Yung

The vast majority of totally enclosed fan cooled (TEFC) squirrel cage induction motors in the 1-to 20-hp range installed in the petroleum and chemical industries are National Electrical Manufacturers Association (NEMA) "T" frames built prior to 1992, NEMA "T" frames built in accordance with the Energy Policy Act of 1992 (commonly referred to as EPAct motors), and the NEMA Premium motors, introduced after the year 2000, that exceed the EPAct efficiency standards. All three types are available in accordance with the IEEE 841 recommended practice and standards. The most obvious difference in these three generations of motors is their efficiency levels. There has been some concern expressed by the users of these motors that to achieve the premium levels of efficiency it was necessary to compromise other performance characteristics and the motor reliability. This article addresses these claims and shows that they are without basis.

  • A.H. Bonnett

The majority of all shaft failures are caused by a combination of various stresses which act upon the rotor assembly. As long as they are kept within the intended design and application limits, shaft failures should not occur during the expected life of the motor. This article provides the reader the methodology to conduct an analysis that will properly identify failures and hopefully take the necessary steps to eliminate them

This paper deals with the electric tracing of the load variation of an induction machine supplied by the mains. A load problem, like a torque dip, affects the machine supply current and consequently it should be possible to use the current pattern to detect features of the torque pattern, using the machine itself as a torque sensor. But current signature depends on many phenomena and misunderstandings are possible. At first the effect of different load anomalies on current spectrum, in comparison with other machine problems like rotor asymmetries, are investigated. Reference is made to low frequency torque disturbances, which cause a quasistationary machine behavior. Simplified relationships, validated by simulation results and by experimental results, are developed to address the current spectrum features. In order to detect on-line anomalies, a current signature extraction is performed by the time-frequency spectrum approach. This method allows the detection of random faults as well. Finally it is shown that a neural network approach can help the torque pattern recognition, improving the interpretation of machine anomalies effects

This paper deals with broken bars detection in induction motors. The hypothesis on which detection is based is that the apparent rotor resistance of an induction motor will increase when a rotor bar breaks. To detect broken bars, measurements of stator voltages and currents are processed by an extended Kalman filter for the speed and rotor resistance simultaneous estimation. In particular, rotor resistance is estimated and compared with its nominal value to detect broken bars. In the proposed extended Kalman filter approach, the state covariance matrix is adequacy weighted leading to a better states estimation dynamic. Its main advantage is the correct rotor resistance estimation even for an unloaded induction motor. As part of this estimation process, it is necessary to compensate for the thermal variation in the rotor resistance. Computer simulations, carried out for a 4 kW four-pole squirrel cage induction motor, provide an encouraging validation of the proposed sensorless broken bars detection technique

In this paper, the effects of inverter harmonics on motor current fault signatures are studied in detail. It is theoretically and experimentally shown that the fault signatures caused by the inverter harmonics are similar and comparable to those generated by the fundamental harmonic on the line current. Theoretically-derived extended relations including bearing fault, eccentricity, and broken rotor bar relations are found to match experimental results. Furthermore, it is observed and reported that the asymmetries on the rotor caused by broken rotor bars increase the amplitude of even harmonics. To confirm these claims, bearing, eccentricity, and broken rotor bar faults are tested and the line current spectrum of each faulty motor is compared with the healthy one. The proposed additional fault data are expected to contribute positively to the inverter-fed motor fault decision making algorithms.

The positive features of neural networks and fuzzy logic are combined together for the detection of stator inter-turn insulation and bearing wear faults in single-phase induction motor. The adaptive neural fuzzy inference systems (ANFISs) are developed for the detection of these two faults. These faults are created experimentally on a single-phase induction motor in the laboratory. The experimental data is generated for the five measurable parameters, viz, motor intakes current, speed, winding temperature, bearing temperature, and the noise of the machine. Earlier, the ANFIS fault detectors are trained for the two input parameters, i.e., speed and current, and the performance is tested. Later, the three remaining parameters are added and the five input ANFIS fault detector is trained and tested. It observed from the simulation results that the five input parameter system predicts more accurate results

  • Mohamed Benbouzid Mohamed Benbouzid

This paper is intended as a tutorial overview of induction motors signature analysis as a medium for fault detection. The purpose is to introduce in a concise manner the fundamental theory, main results, and practical applications of motor signature analysis for the detection and the localization of abnormal electrical and mechanical conditions that indicate, or may lead to, a failure of induction motors. The paper is focused on the so-called motor current signature analysis which utilizes the results of spectral analysis of the stator current. The paper is purposefully written without "state-of-the-art" terminology for the benefit of practising engineers in facilities today who may not be familiar with signal processing

This paper deals with the postfault current control strategies of a five-phase permanent-magnet (PM) motor. The analysis covers both the open circuit of one and two phases and the short circuit at the machine terminal of one phase. The proposed control guarantees safe drive operation after any fault occurrence. For the sake of generality, an analytical model has been used to investigate the properties of each postfault strategy. The results are general, and they apply to PM motor of any power rating. Simulations and experimental results validate the theoretical predictions.

Mechanical rotor imbalances and rotor eccentricities are reflected in electric, electromagnetic, and mechanical quantities. Therefore, many surveillance schemes determine the Fourier spectrum of a single line current in order to monitor the motor condition. Mechanical imbalances give rise to two first-order current harmonics. Due to the interaction of the currents and voltages, both these current harmonics are also reflected by a single harmonic component in the frequency spectrum of the electric power. This single component is easier to assess than both the current harmonics. The technique proposed in this contribution evaluates this imbalance-specific modulation of the electric power. The proposed approach does not determine the Fourier spectrum of a time-domain signal, though. First, the imbalance specific oscillation of the electric power is extracted by a bandpass filter. Then, the averaged pattern of this component is determined by means of an angular data clustering technique. In that way, the oscillation of the electric power in the time domain becomes mapped into a discrete waveform in an angular domain. The amplitude of the fundamental harmonic of these discrete data serves as the imbalance indicator of the proposed scheme. This technique, therefore, overcomes small load and slip fluctuations. Measured results of a mechanically unbalanced machine and a case of combined static and dynamic eccentricity are presented.

  • J.R. Stack
  • Thomas G. Habetler
  • R.G. Harley R.G. Harley

The purpose of this research is to develop a method for experimentally generating in situ bearing faults. To motivate this topic, experimental results are provided that illustrate how the act of removing and replacing test bearings drastically alters the machine vibration and stator current spectral characteristics. Based on this observation, a method is developed that employs an externally applied shaft current to initiate and progress a bearing fault in an accelerated timeframe. This experimental method begins with a new, undamaged bearing and progresses it throughout its entire lifecycle in situ. The test machine is a standard induction motor that can be interfaced with any load and operate at any arbitrary speed or load level throughout the bearing failure process. Data generated by this experimental method can then be used to evaluate the performance of various bearing condition monitoring schemes.

  • A.H. Bonnett

The squirrel-cage induction motor remains the workhorse of the petrochemical industry because of its versatility and ruggedness. However, it has its limitations, which, if exceeded, will cause premature failure of the stator, rotor, bearings or shaft. This paper is the final abridgement and update of six previous papers for the Petroleum and Chemical Industry Committee of the IEEE Industry Applications Society presented over the last 24 years and includes the final piece dealing with shaft failures A methodology is provided that will lead operations personnel to the most likely root causes of failure. Check-off sheets are provided to assist in the orderly collection of data to assist in the analysis. As the petrochemical industry evolves from reactive to time based, to preventive, to trending, to diagnostics, and to a predictive maintenance attitude, more and more attention to root cause analysis will be required. This paper will help provide a platform for the establishment of such an evolution. The product scope includes low- and medium-voltage squirrel-cage induction motors in the 1-3000 hp range with anti friction bearings. However, much of this material is applicable to other types and sizes

  • O.V. Thorsen
  • M. Dalva

This report presents a survey of the reliability of squirrel cage motors on board drilling, production, and other platforms offshore, together with cage motors in the petrochemical industry, gas terminals, and refineries onshore. Most of the activity in this connection is related to The North Sea that offers a tough environment for motors. The collected data have been treated statistically, and the faults sorted according to supply and motor data, driving conditions, electrical protection, maintenance, and so on, and further analyzes failure-initiators, contributors, and underlying causes. Comparisons between this survey and a survey from the IEEE Motor Reliability Working Group [8] have been done. The report also pays some attention to methods for monitoring of machinery and detecting of faults

  • A.H. Bonnett
  • G.C. Soukup

The squirrel cage induction motor has limitations, which, if exceeded, will result in premature failure of the stator or rotor. The authors identify the various causes of stator and rotor failures. A specific methodology is proposed to facilitate an accurate analysis of these failures. Failures of the bearings and lubrication systems are excluded

This paper deals with an application of the Wigner-Ville distribution (WVD) and with usual digital-processing techniques, such as the short-time Fourier transform (STFT), used in dedicated instrumentation for measuring nonstationary signals. The processed real signals are made analytic by means of Hilbert transformations; then suitable implementations of the windowed STFT and of the pseudo Wigner-Ville distribution (PWVD) in the time domain have been performed. Particularly, the fast Hartley transform (FHT) is used to evaluate the PWVD in the real domain. Furthermore, the use of an efficient interpolation algorithm and of a suitable flat-top windowing function is proposed in order to give accurate real-time frequency and amplitude measurements, respectively. With this aim, a dedicated digital system was set up, which uses the LabVIEW software to create virtual instruments (VI), suitable to process the data sequences. Finally, applications of the suggested techniques in analyzing noisy data were also investigated

  • Rajesh Rajamani

This paper presents some fundamental insights into observer design for the class of Lipschitz nonlinear systems. The stability of the nonlinear observer for such systems is not determined purely by the eigenvalues of the linear stability matrix. The correct necessary and sufficient conditions on the stability matrix that ensure asymptotic stability of the observer are presented. These conditions are then reformulated to obtain a sufficient condition for stability in terms of the eigenvalues and the eigenvectors of the linear stability matrix. The eigenvalues have to be located sufficiently far out into the left half-plane, and the eigenvectors also have to be sufficiently well-conditioned for ensuring asymptotic stability. Based on these results, a systematic computational algorithm is then presented for obtaining the observer gain matrix so as to achieve the objective of asymptotic stability

Contribution au diagnostic de la machine asynchrone par estimation paramétrique

  • BACHIR S.

Conception et implémentation d'un Metaé-modèle de machines asynchrones en défaut

  • BAZINE S.

13th European Conference on Power Electronics and Applications EPE

  • BAZINE I.B.A.
  • BAZINE S.
  • TNANI S.
  • CHAMPENOIS G.

Diagnostic - maintenance - disponibilité des machines tournantes

  • BIGRET R.
  • FÉRON J.L.

Recherche de signature électromagnétique des défauts dans une machine asynchrone et synthèse d'observateurs en vue du diagnostic

  • BOUMEGOURA T.

Diagnostic des défauts des machines asynchrones par reconnaissance des formes

  • CASIMIR R.

Convertisseurs et machines électriques

  • GRELLET G.

Vibration produced in squirrel-cage induction motors having broken rotor bars and interbar currents

  • MULLER

Identification et commande des machines électriques

  • LORON L.

Calculpratique des alternateurs et des moteurs asynchrones

  • LOUTZKY S.

Contribution à la modélisation et à l'estimation paramétrique des machines électriques à courant alternatif: Application au diagnostic

  • MOREAU S.

Diagnostic par reconnaissance des formes : Application à un ensemble convertisseur-machine asynchrone

  • ONDEL O.

Diagnostic des machines asynchrones: modèles et outils paramétriques dédiés à la simulation et a la détection de défauts

  • SCHAEFFER E.

Identification des systèmes

  • TRIGEASSOU J.-C
  • POINOT T.