Partial discharge localization in power transformer tanks using machine learning methods.

This paper presents a comparison of machine learning (ML) methods used for three-dimensional localization of partial discharges (PD) in a power transformer tank. The study examines ML and deep learning (DL) methods, ranging from support vector machines (SVM) to more complex approaches like convolutional neural networks (CNN). Multiple case studies are considered, each with different attributes, including sensor position, frequency content of the PD signal, and size of the transformer tank. The paper focuses on predicting the PD location in three-dimensional space using single-sensor electric field measurements. Various aspects of each method are analyzed, such as the input signal, core methodology, correlation coefficient between the predicted location and the actual location, and root mean square error (RMSE). These features are discussed and compared across the different methods. The results indicate that the CNN model exhibits superior performance in terms of location accuracy among the methods considered.

Direct observations of X-rays produced by upward positive lightning.

X-rays have been observed in natural downward cloud-to-ground lightning for over 20 years and in rocket-triggered lightning for slightly less. In both cases, this energetic radiation has been detected during the stepped and dart leader phases of downward negative flashes. More recently, X-rays have also been reported during the dart leader phase of upward negative flashes. In this study, we present the observations of four upward positive lightning flashes from the Säntis Tower (2.5 km ASL) in Switzerland. These consist of the simultaneous records of electric current passing through the tower, and electric field strength and X-ray flux 20 m from the tower base. One of the flashes was captured by a high-speed camera operating at 24,000 frames per second, stills from which are also presented. We detected X-rays during the initial phase of upward negative leader propagation, which can be associated with the leader-stepping process from electric field and current waveforms. To the best of our knowledge, this is the first time that such measurements are reported in the literature. The obtained time-synchronised data confirm that the X-ray emissions detected are associated with the initial steps of the upward negative leader. The frequency and energy of X-ray pulses appear to decrease as functions of time, with pulses disappearing altogether within the first millisecond of the leader initiation. X-ray emission also appears to be correlated with the maximum current-derivative and the electric field change of leader steps, consistent with cold electron runaway. These observations contribute to improving our understanding of upward lightning, which is a primary source of damage to tall structures such as wind turbines and telecommunications towers, as well as aircraft during takeoff and landing.

Time-reversed electromagnetic fields in anisotropic media.

Electromagnetic time reversal is commonly used for field imaging and focusing. This Letter builds upon the concept of the time-reversal cavity, which constitutes the main theoretical framework of time reversal theory. We study the behavior of the fields using modern methods of mathematical physics involving Colombeau generalized functions. This approach allows for a direct expression of time-reversed electric and magnetic fields in anisotropic time-reversal-invariant and nonreciprocal media. Moreover, the results hold for any arbitrary localized source and can readily be applied beyond the dipole approximation. Finally, a general result allows the prediction of the quality of focusing of the time-reversed fields as a function of the electrical permittivity and the magnetic permeability tensors in homogeneous anisotropic media, which contributes to the understanding of time reversal in complex media such as super-resolution enabling metamaterials.

The time-domain Cartesian multipole expansion of electromagnetic fields.

Time-domain solutions of Maxwell's equations in homogeneous and isotropic media are paramount to studying transient or broadband phenomena. However, analytical solutions are generally unavailable for practical applications, while numerical solutions are computationally intensive and require significant memory. Semi-analytical solutions (e.g., series expansion), such as those provided by the current theoretical framework of the multipole expansion, can be discouraging for practical case studies. This paper shows how sophisticated mathematical tools standard in modern physics can be leveraged to find semi-analytical solutions for arbitrary localized time-varying current distributions thanks to the novel time-domain Cartesian multipole expansion. We present the theory, apply it to a concrete application involving the imaging of an intricate current distribution, verify our results with an existing analytical approach, and compare the proposed method to a finite-difference time-domain numerical simulation. Thanks to the concept of current "pixels" introduced in this paper, we derive time-domain semi-analytical solutions of Maxwell's equations for arbitrary planar geometries.

Laser-guided lightning.

Lightning discharges between charged clouds and the Earth's surface are responsible for considerable damages and casualties. It is therefore important to develop better protection methods in addition to the traditional Franklin rod. Here we present the first demonstration that laser-induced filaments-formed in the sky by short and intense laser pulses-can guide lightning discharges over considerable distances. We believe that this experimental breakthrough will lead to progress in lightning protection and lightning physics. An experimental campaign was conducted on the Säntis mountain in north-eastern Switzerland during the summer of 2021 with a high-repetition-rate terawatt laser. The guiding of an upward negative lightning leader over a distance of 50 m was recorded by two separate high-speed cameras. The guiding of negative lightning leaders by laser filaments was corroborated in three other instances by very-high-frequency interferometric measurements, and the number of X-ray bursts detected during guided lightning events greatly increased. Although this research field has been very active for more than 20 years, this is the first field-result that experimentally demonstrates lightning guided by lasers. This work paves the way for new atmospheric applications of ultrashort lasers and represents an important step forward in the development of a laser based lightning protection for airports, launchpads or large infrastructures.

Secondary Fast Breakdown in Narrow Bipolar Events.

The physical mechanism of Narrow bipolar events (NBEs) has been studied for decades but it still holds many mysteries. Recent observations indicate that the fast breakdown discharges that produce NBEs sometimes contain a secondary fast breakdown that propagates back in the opposite direction but this has not been fully addressed so far in electromagnetic models. In this study, we investigate fast breakdown using different approaches that employ a Modified Transmission Line with Exponential decay (MTLE) model and propose a new model, named "rebounding MTLE model," which reproduces the secondary fast breakdown current in NBEs. The model provides new insights into the physics of the fast breakdown mechanism.

Säntis lightning research facility: a summary of the first ten years and future outlook.

The Säntis Tower was instrumented in May 2010 to measure currents of lightning discharges striking the structure. Since then, the system has been recurrently updated and expanded. Presently, data associated with lightning striking the tower are collected at six different sites. The facility is equipped with a current measurement system, three electric field stations, an electrostatic field mill, two x­ray sensors, a high-speed camera, and four slow cameras. This paper presents the latest measurement configuration at the facility. Other temporarily loaned instruments are also briefly described. Furthermore, examples of some of the data that have been gathered and analyzed are given, and an outlook as well as future plans for the facility are presented.

An experimental validation of partial discharge localization using electromagnetic time reversal.

The localization of partial discharge (PD) sources is of importance for the monitoring and maintenance of power transformers. Time difference of arrival (TDoA) based methods are widely adopted in the literature for the localization of PDs. Recently, time reversal (TR) was suggested as an efficient means to locate PD sources. As opposed to TDoA, which needs at least 4 sensors, TR is able to locate PD sources in power transformers with only one sensor. Moreover, it needs neither line-of-sight wave propagation from the PD sources to the sensor nor time synchronization. In this study, we present for the first time an experimental demonstration of the ability of the TR process to locate PD sources. A typical TR process includes three steps: (1) recording the PD-emitted field by a sensor, (2) time reversing and back injecting the signal into the medium, (3) using a proper criterion to obtain the focusing point which corresponds to the location of the PD source. In this work, we present a laboratory setup in which steps one and two are performed experimentally, both in the frequency and in the time domain. The obtained peak electric field value is used as a criterion in the third step. It is found that the accuracy of the proposed method is better than 2.5 cm in a transformer tank model with dimensions 73 × 73 × 103 cm3. The effects of the presence of scatterers such as transformer windings are also investigated experimentally and found not to affect the location accuracy of the method.

Partial Discharge Localization Using Time Reversal: Application to Power Transformers.

In this work, we present a novel technique to locate partial discharge (PD) sources based on the concept of time reversal. The localization of the PD sources is of interest for numerous applications, including the monitoring of power transformers, Gas Insulated Substations, electric motors, super capacitors, or any other device or system that can suffer from PDs. To the best of the authors' knowledge, this is the first time that the concept of time reversal is applied to localize PD sources. Partial discharges emit both electromagnetic and acoustic waves. The proposed method can be used to localize PD sources using either electromagnetic or acoustic waves. As a proof of concept, we present only the results for the electromagnetic case. The proposed method consists of three general steps: (1) recording of the waves from the PD source(s) via proper sensor(s), (2) the time-reversal and back-propagation of the recorded signal(s) into the medium using numerical simulations, and (3) the localization of focal spots. We demonstrate that, unlike the conventional techniques based on the time difference of arrival, the proposed time reversal method can accurately localize PD sources using only one sensor. As a result, the proposed method is much more cost effective compared to existing techniques. The performance of the proposed method is tested considering practical scenarios in which none of the former developed methods can provide reasonable results. Moreover, the proposed method has the unique advantage of being able to locate multiple simultaneous PD sources and doing so with a single sensor. The efficiency of the method against the variation in the polarization of the PDs, their length, and against environmental noise is also investigated. Finally, the validity of the proposed procedure is tested against experimental observations.

Single-Sensor Source Localization Using Electromagnetic Time Reversal and Deep Transfer Learning: Application to Lightning.

Electromagnetic Time Reversal (EMTR) has been used to locate different types of electromagnetic sources. We propose a novel technique based on the combination of EMTR and Machine Learning (ML) for source localization. We show for the first time that ML techniques can be used in conjunction with EMTR to reduce the required number of sensors to only one for the localization of electromagnetic sources in the presence of scatterers. In the EMTR part, we use 2D-FDTD method to generate 2D profiles of the vertical electric field as RGB images. Next, in the ML part, we take advantage of transfer learning techniques by using the pretrained VGG-19 Convolutional Neural Network (CNN) as the feature extractor tool. To the best of our knowledge, this is the first time that the knowledge of pretrained CNNs is applied to simulation-generated images. We demonstrate the skill of the developed methodology in localizing two kinds of electromagnetic sources, namely RF sources with a bandwidth of 0.1-10 MHz and lightning impulses. For the localization of lightning, based on the experimental recordings in the Säntis region, the new approach enables accurate 2D lightning localization using only one sensor, as opposed to current lightning location systems that need at least two sensors to operate.

The Propagation Effects of Lightning Electromagnetic Fields Over Mountainous Terrain in the Earth-Ionosphere Waveguide.

In this paper, a full-wave two-dimensional Finite-Difference-Time-Domain model is developed to evaluate the propagation effects of lightning electromagnetic fields over mountainous terrain in the Earth-ionosphere waveguide. In the model, we investigate the effect of the Earth-ionosphere waveguide structure and medium parameters, including the effect of the ionospheric cold plasma characteristics, the effect of the Earth curvature, and the propagation effects over mountainous terrain. For the first time, the obtained results are validated against simultaneous experimental data consisting of lightning currents measured at the Säntis Tower and electric fields measured in Neudorf, Austria, located at 380-km distance from the tower. It is shown that both the time delays and amplitudes of the lightning electromagnetic fields at 380-km distance can be strongly affected by the ionospheric electron density profile, the mountainous terrain, and the Earth curvature. After taking into account the effect of the irregular terrain between the Säntis Tower and the field measurement station, the vertical electric fields calculated by using our model are found to be in good agreement with the corresponding measured cases occurred in both daytime and nighttime. The ideal approximation used in either the classical solutions or the simplified models might lead to inaccuracies in the estimated reflection height. Furthermore, we discuss the sensitivity of our results by considering different return stroke models, as well as different typical values of the return stroke speed and of the ground conductivity.