Portal de Periódicos da CAPES

Advances in Fault Diagnostics and Post‐Fault Operation of Electrical Drives

2021; Institution of Engineering and Technology; Volume: 15; Issue: 7 Linguagem: Inglês

10.1049/elp2.12100

1751-8679

V. Ambrožič, António J. Marques Cardoso, Gerasimos Rigatos,

Machine Fault Diagnosis Techniques

Modern electrical drives cover various applications differing in power range, operational demands, complexity, reliability and safety requirements. Following the need for a fast and reliable assessment of the drive's health and deciding on a consequent operational mode, different diagnostic approaches emerged in the last two decades. These methods can diagnose a variety of possible faults in all components of the electrical drive. Electrical machines are susceptible to different types of faults, which mainly depend on their internal structure: from mechanical faults (bearings, gearbox) to electrical (insulation faults), electromechanical (rotor faults), and magnetic faults (demagnetisation of permanent magnets). On the other hand, faults in converter-supplied drives could be caused by power electronics components or sensors' failure. In addition to fault diagnostics, significant advancements have also been made in post-fault operation and fault tolerance approaches. A considerable part of new research is focused on new converters' topologies and multiphase drives. This special issue aims to present the current tendencies in fault diagnostics and post-fault operation of various electrical drives. A total of 74 articles were submitted to this special issue. After a rigorous selection process, 17 articles have been accepted, covering a wide area of electric power applications and stretching over multiple topics. In their article ‘Fault-tolerant control strategy of open-winding brushless doubly-fed wind power generator based on direct power control’, Jin et al. present a study on a system under converter switch fault. The topology of the dual converter fed open-winding BDFRG is adopted to improve post-fault operation capability. For post-fault operation, the appropriate voltage vectors of control winding are selected to directly adjust the power winding's active and reactive power based on the DPC theory. Switching tables are formulated under different switch faults. The article ‘Second-order SMO-based sensorless control of IM drive: experimental investigations of observer sensitivity and system reconfiguration in post-fault operation mode’ (Rebah et al.) presents an experimental study of the robustness of speed-controlled induction motor drives without a speed sensor in the presence of inverter open-switches faults. Accordingly, a super twisting algorithm-based observer is proposed for motor speed estimation. Experimental results on pre-fault, post-fault and regeneration operation modes are performed upon a 3 kW induction motor. The article ‘Frequency analysis in fault detection of dual-channel BLDC motors with combined star-delta winding’ (Korkosz et al.) compares the properties of DCBLDCM with permanent magnets in three configurations of both channels’ stator windings. Each type of winding configuration was evaluated in terms of tolerance to a discontinuity in the channel. In a case study, adopting a 24/10 DCBLDCM, the harmonic spectra of a selected voltage signal with respect to the artificial neutral point are presented. It has been demonstrated that the FFT analysis provides unambiguous information on the operating status of the individual channels. Compared with the traditional star and delta winding configurations, the advantages of the combined star-delta winding are presented in the conclusions. Efficiency enhancement through flux current reduction is one of the possible methods for stable post-fault operation of induction motors within the current boundaries. However, the details of power components, that is achievable torque and speed, are usually not addressed. The article ‘Effects of flux derating methods on torque production of fault-tolerant polyphase induction drives’ (Tousizadeh et al.) studies flux and/or torque current partitioning on the post-fault capability of polyphase induction machines considering achievable torque and speed. It is demonstrated that the magnetising inductance characteristics, which depends on the machine's design and its power rating, have a profound effect on the choice of post-fault current partitioning, and hence machine performance. The article ‘Open-circuit fault-tolerant operation of PMSG drives for wind turbine systems using a computationally efficient model predictive current control’ by Jlassi et al. is focused on improving the reliability and availability levels of wind turbines, in which back-to-back converters are very prone to fail. The proposed reconfigurable converter consists of a five-leg converter with a shared leg that connects the generator's first phase to the grid's three phases. A three-switch three-phase rectifier is adopted to achieve fault tolerance for the PMSG-side converter. In their article ‘Modelling and vector control of dual three-phase PMSM with one phase open’, Hu et al. propose general mathematical modelling for the open-phase machine, accounting for the mutual coupling between two sets of three-phase windings and the second harmonic inductance. Both permanent flux linkages and currents become DC values in a dq-reference frame; therefore, the conventional PI controller can be used for the currents' control. Modified fault-tolerant vector control with or without dedicated feed-forward compensation is employed to validate the proposed modelling. This article ‘Comparative analysis of the operating performance, magnetic field and temperature rise of the three-phase permanent magnet synchronous motor with or without fault-tolerant control under single-phase open-circuit fault’, by Li et al. establishes a system-level calculation model. It consists of a permanent magnet synchronous motor and inverter, including different control strategies under fault. Based on the finite element method, the current fault characteristic and the change of the internal magnetic field distribution before and after the fault are studied. The temperature rise in the motor with and without fault-tolerant control is compared under two common fault conditions, including the quantification of the variation range of magnetic flux density and temperature rise. In the article ‘Open-phase fault-tolerant driving operation of dual inverter based traction drive’, Pathmanathan et al. present an approach that utilises a contactor to connect the switching nodes of the two half-bridges. These are typically used for AC charging in dual-inverter configuration. The half-bridges are modulated to provide a path for zero-sequence current and enhance the available voltage vector space by injecting a controllable common-mode voltage between the two inverters. A control method and a five-level modulator based on zero-sequence voltage control and phase-shifted carriers are introduced to accommodate this novel remedial method. Thus, the open-phase faulted dual inverter drive can operate at twice the speed range of the conventional open-phase fault-tolerant approach. Liu et al. propose a speed-loop frequency-adaptive periodic controller and a current-loop optimal harmonic periodic controller for a fault-tolerant surface permanent magnet synchronous motor drive system, including standard operating conditions and faulty operating conditions (IGBT open- and short-circuit) in the article ‘Speed-loop frequency-adaptive and current-loop optimal harmonic periodic controllers for fault-tolerant SPMSM drive systems'. Experimental results demonstrate that the proposed advanced periodic controllers perform better than the PI controller and the classic periodic controller, including transience, load disturbance, and tracking responses under normal and faulty conditions. ‘Static and dynamic eccentricity fault diagnosis of large salient pole synchronous generators by means of external magnetic field’ is addressed by Ehya et al. The field is measured by two search coils installed on the backside of the stator yoke. Advanced signal processing tools utilising wavelet entropy analyse the induced electromotive force in search coils to extract the fault index. The proposed index does not require the threshold to recognise the incipient fault. It is sensitive to a low degree of fault while being robust to load variation and noise that may generate a false alarm. Although the popularity of stray and air-gap flux monitoring methods is increasing, using magnetic flux for mechanical faults' detection has not drawn much attention. At the same time, vibration analysis continues to be popular in the industry. The article ‘Detection of simultaneous mechanical faults in 6 kV pumping induction motors using combined MCSA and stray flux methods’ (Gyftakis et al.) comes to bridge this gap via detecting mechanical faults of 6 kV induction motors in a pumping station. The diagnostic procedure mainly involves the stator current, stray flux monitoring, and harmonic index analysis. The localisation of the fault has been made possible via oscillometer readings. It is demonstrated that mechanical faults have a very different impact on the stator current and the flux signals, while the flux is not sensitive to the bearing fault mechanisms. Convolutional neural networks (CNNs) have redefined the state-of-the-art accuracy for bearing fault detection and identification, extracting location invariant feature mappings without the need for prior expert knowledge. Karnavas et al. propose a deep learning (DL) model that concatenates the features produced from two neural streams in the article ‘Extracting spatially global and local attentive features for rolling bearing fault diagnosis in electrical machines using attention stream networks’. Each of them consists of an attention mechanism that intends to learn different representations of the input vector. Finally, it produces a feature mapping that contains global and spatial local information. The proposed DL model achieves 99.60% in the Case Western Reserve University bearing dataset and 99.10% in the Paderborn University bearing dataset. In ‘Eccentricity fault detection in brushless doubly-fed induction machines’, Afshar et al. address the issue of static, dynamic, and mixed eccentricity. The authors propose a novel fault detection method based on motor current signal analysis to determine stator current harmonics, induced by the nested-loop rotor slot harmonics as fault indices, under healthy conditions and with different types of rotor eccentricity. Analytical winding function approach, finite element analysis, and experimental tests on a prototype D180 BDFIM are used in this study to validate the proposed fault detection technique. In the article ‘Robustification of fault detection algorithm in a three-phase induction motor using MCSA for various single and multiple faults’, Kompella et al. perform a critical comparison between two popular pre-fault component cancellation techniques. A reliability test is implemented to develop a robust algorithm for fault diagnosis in a three-phase induction motor. Various single and multiple faults experienced by the induction motor are created and tested to examine the algorithm's effectiveness by repeating the process and testing for consistency on a three phase, 1.5 kW motor. Various feature extraction parameters are computed and compared to identify the best estimate of the fault and its severity. In ‘Application of simplified convolutional neural networks for initial stator winding fault detection of PMSM drive using different raw signal data’, Skowron et al. present a method applying direct signal analysis and a CNN. The CNN structures were optimised to constitute a balance between the high efficiency of fault detection and a small number of network parameters. Stator phase currents, phase-to-phase voltages and axial flux signals are tested as CNN inputs without preliminary pre-processing. The article aims to show the possibility of detecting the incipient shorted turns in the stator winding based on the information obtained directly from the measured signals and indicate the influence of the drive's diagnostic quantities and operating condition on the neural network structure. The work ‘A comprehensive approach to convolutional neural networks based condition monitoring of permanent magnet synchronous motor drives”’ (Pasqualotto and Zigliotto) is devoted to the application of a special kind of neural networks to interpret the data from motor currents for diagnosing demagnetisation and inter-turn fault. The innovation is in the overall approach to neural network training, which does not call for a large set of faulty motors. The result is a comprehensive and effective motor-condition monitoring algorithm, whose hearth is a convolutionary neural network trained by a safe and cheap simulation-based dataset. The generality of the proposed method also paves the way for the detection of other failures and the application to different electrical motors. Typically, a switched reluctance drive requires one current sensor for each phase winding for precise and stable control performance. Failure of any of these sensors inevitably degrades the system performance. Ali and Gao, in ‘A simple current sensor fault-tolerant control strategy for switched reluctance motors in high-reliability applications’, investigate the behaviour of an SRM drive under sensor failure and propose a post-fault control technique by changing the current sensor installation scheme without adding extra sensors. Detailed analysis under steady-state and dynamic conditions has been carried out on a three-phase 12/8 SRM through simulations and experiments. Vanja Ambrožič received the BSc, MSc and PhD degrees from the University of Ljubljana, Slovenia, where he is currently working as a full professor at the Department of Mechatronics, Laboratory of Control Engineering and Power Electronics. He was a longtime head of the Department of Mechatronics. He has authored or co-authored more than 170 scientific papers in various journals and conferences, with four university textbooks on electrical drives, embedded microprocessors, and programmable logic controllers. He has held several invited lectures at foreign universities and conferences. He is also an Editor-at-Large of IET Power Electronics and Associate Editor of IET Electric Power Applications and IEEE Transactions on Industrial Electronics. His main research interests include control of electrical drives and power electronics and diagnostics. Antonio J. Marques Cardoso received the Dipl. Eng., Dr. Eng., and Habilitation degrees from the University of Coimbra, Portugal in 1985, 1995 and 2008, respectively, all in electrical engineering. He was with the University of Coimbra from 1985 to 2011, where he was Director of the Electrical Machines Laboratory. Since 2011, he has been with the University of Beira Interior (UBI), Portugal, where he is full professor at the Department of Electromechanical Engineering and Director of CISE - Electromechatronic Systems Research Centre (http://cise.ubi.pt). He was Vice-Rector of UBI (2013–2014). His current research interests are in fault diagnosis and fault tolerance of electrical machines, power electronics and drives. He is the author of a book entitled Fault Diagnosis in Three-Phase Induction Motors (Coimbra, Portugal: Coimbra Editora, 1991, in Portuguese), editor of a book entitled Diagnosis and Fault Tolerance of Electrical Machines, Power Electronics and Drives (IET/SciTech, UK, 2018), and also the author of around 500 papers published in technical journals and conference proceedings. Gerasimos Rigatos obtained a diploma (1995) and a PhD (2000) both from the Department of Electrical and Computer Engineering, of the National Technical University of Athens (NTUA), Greece. He was a post-doctoral researcher at the Institut de Recherche en Informatique et Systèmes Aléatoires IRISA, France in 2001. Since 2002, he has held a researcher position (currently Grade A—Research Director) at the Industrial Systems Institute (Athena Research Centre on Innovation and Information Technologies), Greece on the topic of ‘Modelling and Control of Industrial Systems’. In 2007, he was an invited professor (maître des conférences) at Université Paris XI (Institut d' Electronique Fondamentale), France. In 2012, he held a lecturer position at the Department of Engineering, Harper-Adams University College, UK on ‘Mechatronics and Artificial Intelligence’. He has also been an adjunct professor in Greek Universities, where he has taught courses on systems and control theory. His research interests include the areas of control and robotics, computational intelligence and adaptive systems, mechatronics, optimisation and fault diagnosis. He holds an Editor position in accredited international journals. He is a senior member of IEEE, a member and CEng of IET, and also a member of IMACS. He held a 4-year visiting professor position at the University of Salerno, Italy, between 2016 and 2020, and between 2016 and 2019, he held a 3-year visiting professor position at the University of Northumbria, UK. During 2021–2022, he was given a visiting professor's position at Ecole Centrale de Nantes, France.