A Hybrid Sensor Fault Diagnosis for Maintenance in Railway Traction Drives

Abstract

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.Sensors 2020, 20, 962 2 of 21 which normally requires an important effort in complex systems, but as an advantage, it is suitable for on-board and real-time implementation, as it requires neither a large quantity of data, nor high computational resources. The model-based techniques can be divided into two parts: the residual generation and the residual evaluation. On the other hand, data-driven techniques do not need a knowledge of the physical system, but the computational burden is normally higher. Moreover, data quantity to process and storage is large, so it makes difficult its on-board implementation.

Data-Driven Fault Diagnosis for Electric Drives: A Review

“…It is worth mentioning that the information redundancy added to AI techniques make possible to build a complete CM platform, from data acquisition (DA) to prognostics assessment (PA) levels. • Hybrid methodology: As the aforementioned active supervision techniques have their pros and cons, a new trend has emerged, which tries to integrate together these complementary methods in order to achieve a better performance [41,42]. It is commonly agreed that hybrid schemes would provide better solutions to a complex system.…”

New Business Models from Prescriptive Maintenance Strategies Aligned with Sustainable Development Goals

“…Embedded sensors, software, controllers, connectivity, breakthroughs in networking, edge and fog computing, cloud technology, faster CPUs, cheaper memory, energy efficiency, and miniaturisation are all converging to create low-cost processing power and data storage everywhere and they became the key enabling technologies fuelling up the evolution in the paradigms [32][33][34]. As presented in Merluzzi and Brunetti [35] the "Reliability-Centered Maintenance" represents a systematic approach to evaluate a facility's equipment and resources to best mate the two and result in a high degree of facility reliability and costeffectiveness.…”

Application of the MEMS Accelerometer as the Position Sensor in Linear Electrohydraulic Drive

“…Based on the literature review, it can be concluded that a MEMS-type acceleration sensor has not been used so far in the way presented by the author for positioning the electrohydraulic drive. The authors of the above-mentioned papers focused mainly on the use of this type of sensor for diagnostics of various types of hydraulic systems, examining parameters such as vibrations [ 19 , 20 , 21 , 22 ]. Other popular applications are measurements of changes in angular position, where the signal from the acceleration sensor is combined with the signal from a gyroscope, among others.…”