Optical fibre based quench detection in HTS applications using machine learning classifiers

A Mach-Zehnder Interferometer (MZI) based optical fibre sensing technique, developed and patented by EPFL, is an efficient and economical way to detect hotspots in High Temperature Superconductor (HTS) applications. Due to the MZI sensitivity being a composite of strain sensitive and temperature sensitive contributions, the MZI gives an instantaneous response to a quench (within 10 ms), because of the quick strain transfer to the optical fibre. However, the MZI output signal also manifests the environmental noise caused by mechanical vibrations, bubbling in the cryostat and temperature variations, along with the response to the quench. This presents the problems of false alarms and indiscernible response to a quench. Discrete wavelet transform (DWT) has been proven to be a useful tool for feature extraction in different fields requiring signal categorization and hence holds the potential to enable quench recognition in the MZI output. This paper proposes an effective approach of performing DWT based feature extraction on experimental data and subsequently using the extracted features for the MZI response classification using two machine learning based classification techniques: k-nearest neighbours (KNN) and Artificial Neural Network (ANN). For this manuscript, experiments were performed using MZI for quench detection in an HTS tape. Feature extraction was then implemented on these experimental measurements using discrete wavelet coefficients extracted at different decomposition levels from the MZI output; these features were then used to train the KNN and ANN models for identifying quench in the MZI signal. This method could be a valuable supplement to the MZI technique by enabling the development of a real time application that can process the MZI output data as well as eliminate the occurrences of false alarms; thereby facilitating reliable quench detection. With this development, the MZI technique would become an even more attractive solution for the health monitoring of HTS applications.

A fibre optic and machine learning based health monitoring system for superconducting power applications in modern supergrids

Arooj Akbar

The concerns for climate change and the transition to renewables is increasing the relevance of superconducting applications: High Temperature Superconductor (HTS) cables for efficient power transmission, superconducting magnets for nuclear fusion, superconducting fault current limiters (SFCLs) for HVDC grid protection and superconducting compact generators for wind turbines, to name a few. However, to tap into the full potential of HTS applications, the main problem of hotspots needs to be addressed. Hotspots are localized points of heating, which can arise in HTS devices mainly due to the intrinsic inhomogeneity of critical current along the superconductor length. The motivation behind this thesis, was to develop a health monitoring technique for a SFCL device under the European union project FastGrid. This device required a fast hotspot detection time of 10 ms in order to timely open the DC breaker in order to protect the device. During the course of this thesis an extremely fast and economical optical fibre sensing based hotspot detection technique has been developed and patented. The technique uses the Mach-Zehnder Interferometer (MZI) and is an efficient and economical way to detect hotspots in HTS applications. Due to the MZI sensitivity being a composite of strain sensitive and temperature sensitive contributions, the MZI gives an instantaneous response to a hotspot (within 10 ms), because of the quick strain transfer to the optical fibre. The research carried out under this thesis has involved experimentation on HTS tapes (10 cm to 1 m in length). The technique has also been integrated and tested with a SFCL pancake prototype with long lengths of conductor (12 m and 17 m). The thesis has also focused on finite element modelling to better understand the MZI sensitivity. The experimentation and modelling together, has enabled a better understanding of the MZI response to hotspots and the challenges of using MZI for HTS health monitoring. Due to the extremely high sensitivity of the MZI , the MZI output signal can also manifest the environmental noise caused by mechanical vibrations, bubbling in the cryostat and temperature variations, along with the response to heating in the sample. This presents the problems of false alarms and indiscernible response to hotspots. The work done for this thesis has also focused substantially on finding a suitable data analysis technique to supplement the MZI method. A discrete wavelet transform (DWT) based feature extraction together with a machine learning based classification for reliable quench detection has been developed in the course of this work with very encouraging results . This is an important development for the MZI technique as it significantly impacts and improves the MZI reliability and practicality for HTS applications. Additionally, this is also of substantial importance to the progress of HTS applications in the industry, especially the power sector which continues to evolve with the shift towards renewables and will rely more on HTS applications like SFCLs and HTS cables. With a reliable hotspot detection technique like the MZI, we can facilitate mass adoption of such HTS applications.

EPFL2022

Adaptive wake and sleep detection for wearable systems

Walter Karlen

Sleep problems and disorders have a serious impact on human health and wellbeing. The rising costs for treating sleep-related chronic diseases in industrialized countries demands efficient prevention. Low-cost, wearable sleep / wake detection systems which give feedback on the wearer's "sleep performance" are a promising approach to reduce the risk of developing serious sleep disorders and fatigue. Not all bio-medical signals that are useful for sleep / wake discrimination can be easily recorded with wearable systems. Sensors often need to be placed in an obtrusive location on the body or cannot be efficiently embedded into a wearable frame. Furthermore, wearable systems have limited computational and energetic resources, which restrict the choice of sensors and algorithms for online processing and classification. Since wearable systems are used outside the laboratory, the recorded signals tend to be corrupted with additional noise that influences the precision of classification algorithms. In this thesis we present the research on a wearable sleep / wake classifier system that relies on cardiorespiratory (ECG and respiratory effort) and activity recordings and that works autonomously with minimal user interaction. This research included the selection of optimal signals and sensors, the development of a custom-tailored hardware demonstrator with embedded classification algorithms, and the realization of experiments in real-world environments for the customization and validation of the system. The processing and classification of the signals were based on Fourier transformations and artificial neural networks that are efficiently implementable into digital signal controllers. Literature analysis and empiric measurements revealed that cardiorespiratory signals are more promising for a wearable sleep / wake classification than clinically used signals such as brain potentials. The experiments conducted during this thesis showed that inter-subject differences within the recorded physiological signals make it difficult to design a sleep / wake classification model that can generalize to a group of subjects. This problem was addressed in two ways: First by adding features from another signal to the classifier, that is, measuring the behavioral quiescence during sleep using accelerometers. Conducted research on different feature extraction methods from accelerometer data showed that this data generalizes well for distinct subjects in the study group. In addition, research on user-adaptation methods was conducted. Behavioral sleep and wake measures, notably the measurement of reactivity and activity, were developed to build up a priori knowledge that was used to adapt the classification algorithm automatically to new situations. This thesis demonstrates the design and development of a low-cost, wearable hardware and embedded software for on-line sleep / wake discrimination. The proposed automatic user-adaptive classifier is advantageous compared to previously suggested classification methods that generalize over multiple subjects, because it can take changes in the wearer's physiology and sleep / wake behavior into account without adjustment from a human expert. The results of this thesis contribute to the development of smart, wearable, bio-physiological monitoring systems which require a high degree of autonomy and have only low computational resources available. We believe that the proposed sleep / wake classification system is a first promising step toward a context-aware system for sleep management, sleep disorder prevention, and reduction of fatigue.

EPFL2009

Modeling and Characterization of Coated Conductors Applied to the Design of Superconducting Fault Current Limiters

François Roy

The Superconducting Fault Current Limiter (SFCL) appears to be a device of great interest to efficiently build the electrical grid of tomorrow. With the recent progress made by the superconducting wires manufacturers, there are needs coming from the industry to evaluate the potential of such devices. In the present thesis work, the behavior under external field and transport current of the last generation of wire is investigated. This study is conduct both experimentally and numerically in order to link the physics occurring at the wires level to the design of SFCLs as a whole. From the nature of the material, the resistance appears non-uniformly in high temperature superconductors. For the purpose of building SFCLs it is important to obtain a fast and uniform resistive transition (quench) when a fault occurs through those conductors. This in order to reduce the local heat generation that may damage the device. This fast quenching property is related to the Normal Zone Propagation Velocity (NZPV). In this work the NZPV is measured using a localized magnetic field to initiate quenches in commercial coated conductors. Those velocities have been measured to be larger than 14 cm/s for pulsed currents above the critical value. The NZPV experiments have demonstrated that the superconductor non-uniformity (generated by the localized field) helps to reduce the initial delay before the quench initiation for transport currents in the range of the critical value. However, for larger transport currents the effect of the non-uniformity on the delay is less important since, with increasing transport current amplitudes, the normal state transition has shown to occur more as a consequence of the heat generated in the stabilizer than as the unique consequence of the advancement of the normal zone in the superconductor. From the experimental measurements, it has been shown that a reduction of the liquid-nitrogen temperature (subcoooled) increases the NZPV. This effect has been observed taking into account of the increase of the critical current associated with the temperature reduction. Nevertheless, it is not clear if it is the heat transfer or the estimation of the critical current that is responsible for this effect. In order to validate the numerical models, time-resolved voltage traces obtained from the experiments have been compared to the outputs of the models. Those are based on the thermal- and electrical-diffusion equations. From the simulations, it has been demonstrated that the NZPV can be increased by three methods: by using a thick diffusive substrate, by inserting a resistive interface between the superconductor and the stabilizer as well as by increasing the heat generation in the stabilizer. In light of those results, it seems that the insertion of a resistive layer is the most promising approach to improve the NZPV in coated conductors. As a matter of fact, a resistive interface increases the normal-zone size and keeps an acceptable temperature level along the conductor during quenches. The present work allowed to simulate the flux-flow regime in coated conductors. Comparing those simulations to experimental data have shown that the power-law may be inappropriate to simulate this regime under weak external magnetic fields. In addition, it appears that the role of transient heat transfer with the surroundings needs to be studied in more details to determine the specifications of a prospected SFCL made of coated conductors.

EPFL2010