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.

In this paper, we propose and implement a novel technique to locate Electromagnetic Interference (EMI) sources using the concept of time reversal cavity. We show in an intuitive manner that reflections from the surfaces of a cavity can emulate an infinite number of sensors in the Time Reversal (TR) method. In order to locate EMI sources, the equipment under test (EUT) is placed in a rectangular metallic cavity, of suitable dimensions according to the considered frequency band and the EUT size, equipped with a simple monopole or dipole antenna. We demonstrate that, by using only one sensor, we are able to locate EMI sources by taking advantage of the focusing properties of a time reversal cavity. The entropy criterion is applied to obtain the focusing time slice in which the maximum electric field determines the location of the EMI source. Both two- and three-dimensional numerical simulation schemes are deployed to demonstrate the ability of the proposed technique. The validity of numerical simulations is tested against frequency domain measurements. Compared to conventional EMI tests in anechoic chambers and scanning methods, the proposed technique represents a simpler and cost-effective test method requiring only one sensor (a monopole or dipole antenna).

We present a study on the characteristics of current and electric field pulses associated with upward lightning flashes initiated from the instrumented Säntis Tower in Switzerland. The electric field was measured 15 km from the tower. Upward flashes always begin with the initial stage composed of the upward‐leader phase and the initial‐continuous‐current (ICC) phase. Four types of current pulses are identified and analyzed in the paper: (1) return‐stroke pulses, which occur after the extinction of the ICC and are preceded by essentially no‐current time intervals; (2) mixed‐mode ICC pulses, defined as fast pulses superimposed on the ICC, which have characteristics very similar to those of return strokes and are believed to be associated with the reactivation of a decayed branch or the connection of a newly created channel to the ICC‐carrying channel at relatively small junction heights; (3) “classical” M‐component pulses superimposed on the continuing current following some return strokes; and (4) M‐component‐type ICC pulses, presumably associated with the reactivation of a decayed branch or the connection of a newly created channel to the ICC‐carrying channel at relatively large junction heights. We consider a data set consisting of 9 return‐stroke pulses, 70 mixed‐mode ICC pulses, 11 classical M‐component pulses, and 19 M‐component‐type ICC pulses (a total of 109 pulses). The salient characteristics of the current and field waveforms are analyzed. A new criterion is proposed to distinguish between mixed‐mode and M‐component‐type pulses, which is based on the current waveform features. The characteristics of M‐component‐type pulses during the initial stage are found to be similar to those of classical M‐component pulses occurring during the continuing current after some return strokes. It is also found that about 41% of mixed‐mode ICC pulses were preceded by microsecond‐scale pulses occurring in electric field records some hundreds of microseconds prior to the onset of the current, very similar to microsecond‐scale electric field pulses observed for M‐component‐type ICC pulses and which can be attributed to the junction of an in‐cloud leader channel to the current‐carrying channel to ground. Classical M‐component pulses and M‐component‐type ICC pulses tend to have larger risetimes ranging from 6.3 to 430 μs. On the other hand, return‐stroke pulses and mixed‐mode ICC pulses have current risetimes ranging from 0.5 to 28 μs. Finally, our data suggest that the 8‐μs criterion for the current risetime proposed by Flache et al. is a reasonable tool to distinguish between return strokes and classical M‐components. However, mixed‐mode ICC pulses superimposed on the ICC can sometimes have considerably longer risetimes, up to about 28 μs, as observed in this study.

In this study, first, a comparison on the application of electromagnetic time reversal (EMTR) and time difference of arrival (TDoA) in partial discharge localization in power transformers is presented. A two-dimensional finite-difference time-domain simulation is used to calculate the signal recorded by the sensors. Results show that, in a transformer tank excluding its windings, both methods yield similar results in terms of location accuracy, although the EMTR method only needs one sensor to localize the partial discharge (PD) source while the TDoA method needs at least three sensors in the 2D localization problem. However, the presence of transformer windings leads to a degradation of the performance of the TDoA method if the line of sight from the source to the sensor is blocked by any of the winding blocks. On the other hand, the presence of the transformer windings has an effect on the localization of PD sources that occur between two adjacent phase windings when the distance between the outer winding distances is shorter than the minimum wavelength, λmin. The degradation is directly caused by the diffraction limit. It is shown that, if the distance between two adjacent phase windings is greater than λmin, the EMTR process can locate PD sources occurring between two adjacent phase windings with acceptable accuracy. A case of occurrence of PDs in close proximity (less than λmin/2) to a single metallic object is analyzed both numerically and experimentally. The analysis reveals that although a degradation in the accuracy of the localization is observed compared to the case of longer distances between the PD source and the metallic object, a reasonable localization error of 10 mm (corresponding to λmin/10) is obtained.

We implement an electromagnetic time reversal technique (EMTR) to locate lightning return strokes. The two-dimensional finite difference time domain is employed to simulate the EMTR process in both, the forward-time and the backward-time phases. Scatterers are included in the computational domain to emulate the presence of objects. Three possible criteria to find the optimum time slice of the EMTR process that includes the maximum peak field, maximum peak energy, and last local minimum of entropy are tested and it is found that only the entropy criterion can successfully locate the lightning discharge. Our analysis shows that the EMTR process in both, using an unchanged and a simplified medium for the backward time works reasonably well even with only two sensors. Furthermore, we validated the proposed method via experimental results using waveforms recorded at two sensors at distances of 14.7 and 380 km from the Säntis Tower. The results demonstrate that the EMTR back-propagation process leads to a refocusing of the radiated energy at the location of the Säntis Tower. The ambiguity in the obtained location when only two sensors are used can be resolved either by using an additional sensor or through a more accurate modeling of the terrain.

In this article, we present a machine learning-based method to locate lightning flashes using calculations of lightning-induced voltages on a transmission line. The proposed approach takes advantage of the preinstalled voltage measurement systems on power transmission lines to get the data. Hence, it does not require the installation of additional sensors such as extremely low frequency, very low frequency, or very high frequency. The proposed model is shown to yield reasonable accuracy in estimating two-dimensional geolocations for lightning strike points for different grid sizes up to 100 × 100 km2. The algorithm is shown to be robust against the distance between the voltage sensors, lightning peak current, lightning current rise time, and signal to noise ratio of the input signals.