The need to manufacture more competitive equipment, together with the emergence of the digital technologies from the so-called Industry 4.0, have changed many paradigms of the industrial sector. Presently, the trend has shifted to massively acquire operational data, which can be processed to extract really valuable information with the help of Machine Learning or Deep Learning techniques. As a result, classical Condition Monitoring methodologies, such as model- and signal-based ones are being overcome by data-driven approaches. Therefore, the current paper provides a review of these data-driven active supervision strategies implemented in electric drives for fault detection and diagnosis (FDD). Hence, first, an overview of the main FDD methods is presented. Then, some basic guidelines to implement the Machine Learning workflow on which most data-driven strategies are based, are explained. In addition, finally, the review of scientific articles related to the topic is provided, together with a discussion which tries to identify the main research gaps and opportunities.

During the last decade, due to the increasing importance of reliability and availability, railway industry is making greater use of fault diagnosis approaches for early fault detection, as well as Condition-based maintenance frameworks. Due to the influence of traction drive in the railway system availability, several research works have been focused on Fault Diagnosis for Railway traction drives. Fault diagnosis approaches have been applied to electric machines, sensors and power electronics. Furthermore, Condition-based maintenance framework seems to reduce corrective and Time-based maintenance works in Railway Systems. However, there is not any publication that summarizes all the research works carried out in Fault diagnosis and Condition-based Maintenance frameworks for Railway Traction Drives. Thus, this review presents the development of Health Assessment and Fault Diagnosis in Railway Traction Drives during the last decade.

Due to the increasing importance of reliability and availability of electric traction drives in Railway applications, early detection of faults has become an important key for Railway traction drive manufacturers. Sensor faults are important sources of failures. Among the different fault diagnosis approaches, in this article an integral diagnosis strategy for sensors in traction drives is presented. Such strategy is composed of an observer-based approach for direct current (DC)-link voltage and catenary current sensors, a frequency analysis approach for motor current phase sensors and a hardware redundancy solution for speed sensors. None of them requires any hardware change requirement in the actual traction drive. All the fault detection and isolation approaches have been validated in a Hardware-in-the-loop platform comprising a Real Time Simulator and a commercial Traction Control Unit for a tram. In comparison to safety-critical systems in Aerospace applications, Railway applications do not need instantaneous detection, and the diagnosis is validated in a short time period for reliable decision. Combining the different approaches and existing hardware redundancy, an integral fault diagnosis solution is provided, to detect and isolate faults in all the sensors installed in the traction drive.

Due to the importance of sensors in railway traction drives availability, sensor fault diagnosis has become a key point in order to move from preventive maintenance to condition-based maintenance. Most research works are limited to sensor fault detection and isolation, but only a few of them analyze the types of sensor faults, such as offset or gain, with the aim of reconfiguring the sensor in order to implement a fault tolerant system. This article is based on a fusion of model-based and data-driven techniques. First, an observer-based approach, using a Sliding Mode observer, is utilized for sensor fault reconstruction in real time. Then, once the fault is detected, a time window of sensor measurements and sensor fault reconstruction is sent to the remote maintenance center for fault evaluation. Finally, an offline processing is carried out to discriminate between gain and offset sensor faults, in order to get a maintenance decision-making to reconfigure the sensor during the next train stop. Fault classification is done by means of histograms and statistics. The technique here proposed is applied to the DC-link voltage sensor in a railway traction drive and is validated in a hardware-in-the-loop platform.

The modulation method in inverter-fed ac-drive applications has a great influence in the performance of the entire system as it generates losses in the converter and the machine, degrades the signal waveform quality and causes additional noise and vibrations. By carefully selecting the modulation method, some of these unwanted effects can be improved. This paper explores the influence of some of the most known PWM-based modulation methods as well as the switching frequency selection on the system performance and characteristics: converter and machine losses, harmonic effects or implementation issues. Thanks to the analytical and multidomain simulation-based study, a full comparison of different modulation schemes is provided.

Eddy current losses generated in the rotor of the permanent magnet machines may lead to an excessive magnet heating. This can cause full demagnetization of the magnets so it is very important to analyze the losses on them to adopt solutions to reduce eddy current losses. The most common technique is the magnet circumferential segmentation. The segmentation of the magnetic poles usually is the same for each magnet segment. In this paper a new concept named nonuniform magnet segmentation is presented, where the length of each segment of a pole is chosen to minimize the losses. This method is tested in three different machines with particular characteristics and the results confirm that fewer losses can be achieved with this segmentation technique.