Artificial Generation of Partial Discharge Sources Through an Algorithm Based on Deep Convolutional Generative Adversarial Networks

Abstract

The measurement of partial discharges (PD) in electrical equipment or machines subjected to high voltage can be considered as one of the most important indicators when assessing the state of an insulation system. One of the main challenges in monitoring these degradation phenomena is to adequately measure a statistically significant number of signals from each of the sources acting on the asset under test. However, in industrial environments the presence of large amplitude noise sources or the simultaneous presence of multiple PD sources may limit the acquisition of the signals and therefore the final diagnosis of the equipment status may not be the most accurate. Although different procedures and separation and identification techniques have been implemented with very good results, not having a significant number of PD pulses associated with each source can limit the effectiveness of these procedures. Based on the above, this research proposes a new algorithm of artificial generation of PD based on a Deep Convolutional Generative Adversarial Networks (DCGAN) architecture which allows artificially generating different sources of PD from a small group of real PD, in order to complement those sources that during the measurement were poorly represented in terms of signals. According to the results obtained in different experiments, the temporal and spectral behavior of artificially generated PD sources proved to be similar to that of real experimentally obtained sources. INDEX TERMS Partial discharge, electrical noise sources, machine learning, spectral power clustering technique, clustering.

Composite Insulator Defect Identification Method Based on Acoustic–Electric Feature Fusion and MMSAE Network

“…By smearing carbon powder on the surface of the high-voltage umbrella skirt, the leakage tracking defect caused by a flashover on the surface of the umbrella skirt can be simulated. According to the failure mechanism of insulators [29][30][31][32], the following six kinds of defects are simulated, as shown in Table 1. Figure 2 depicts the actual simulation of various defects.…”

Switchgear Digitalization—Research Path, Status, and Future Work

“…To put it in context: Continuous measurements for temperature monitoring of switchgears are rare or basically non-existent during their entire operation time, and the breaker switching operation is infrequently performed, which makes it extremely difficult to collect data or detect patterns for training algorithms. One solution to the aforementioned problem is the use of a generative adversarial network (GAN), a deep learning algorithm that enables the generation of more data that mimics real data from a limited set of obtained real/experimental data, as in [ 69 ]. A full summary of the integration of condition monitoring and machine learning algorithms for predictive maintenance is presented using medium voltage switchgear as an example in [ 70 ].…”

Partial Discharge Identification in MV Switchgear Using Scalogram Representations and Convolutional AutoEncoder

“…The construction of a large dataset is important to train a deep neural network and avoid bias. Usually this data is not available because of the complexity in performing repetitive tests and some authors [17]- [19] have used generative models [20] to create more data and complete their database.…”