Artificial intelligence-based solutions and applications have great potential in various fields of electrical power engineering. The problem of the electrical reliability of power equipment directly refers to the immunity of high-voltage (HV) insulation systems to operating stresses, overvoltages and other stresses—in particular, those involving strong electric fields. Therefore, tracing material degradation processes in insulation systems requires dedicated diagnostics; one of the most reliable quality indicators of high-voltage insulation systems is partial discharge (PD) measurement. In this paper, an example of the application of a neural network to partial discharge images is presented, which is based on the convolutional neural network (CNN) architecture, and used to recognize the stages of the aging of high-voltage electrical insulation based on PD images. Partial discharge images refer to phase-resolved patterns revealing various discharge stages and forms. The test specimens were aged under high electric stress, and the measurement results were saved continuously within a predefined time period. The four distinguishable classes of the electrical insulation degradation process were defined, mimicking the changes that occurred within the electrical insulation in the specimens (i.e., start, middle, end and noise/disturbance), with the goal of properly recognizing these stages in the untrained image samples. The results reflect the exemplary performance of the CNN and its resilience to manipulations of the network architecture and values of the hyperparameters. Convolutional neural networks seem to be a promising component of future autonomous PD expert systems.

Understanding a partial discharge mechanism at direct current (DC) is an actual research topic that requires both modeling, simulations and measurements. This paper describes an influence of insulating material properties on partial discharges at DC voltage. Modifications of the traditional model reflecting the conditions of partial discharges (PD) inception and post discharge processes at DC voltage have been proposed. The aim was to show the partial discharge mechanisms and draw attention to the role of parameters of insulation materials adjacent to the cavity at DC voltage. The investigations were performed on two kinds of dielectric material used in power cables. Various combinations of specimens were designed to reveal the effect of the material resistivity on the PD activity. Key observations referred to the impact of the void adjacent material resistance on the partial discharge inception voltage threshold at DC voltage. The modified PD model was applied to analyze both inception and post discharge recovery stage. The role of dielectric properties of material adjacent to the void was investigated, highlighting its impact during static inception stage and in charging stage. Despite many simplifications introduced in the model, measurement results have confirmed the role of the dielectric material surrounding the void on partial discharge dynamics. The average time interval between PD pulses revealed a systematic relationship with respect to the applied voltage and specimen resistivity. This value can be considered in the future research for diagnostic indicator at DC voltage.

This paper describes a novel approach to understanding partial discharge (PD) dynamics and mapping of the discharge channels in a void. During long term PD exposure, discharges develop inside channels within a void, leading to a variation in the discharge area on the void wall as well as a clustering effect. In this paper, the development of a method for identifying localized changes in the dielectric structure, through the detection of dispersed channels with reduced resistivity which occur due to multichannel discharges, is described. Discharge spots in selected ranges of surface resistivity, occurring due to the energetic impact of PD, are mapped. This mapping shows both the distribution of local centers on the surface of the specimen, characterized by a reduced value of surface resistivity. Furthermore, during aging tests this surface resistivity exhibited a progressive reduction by four orders of magnitude, justifying the occurrence of an effective discharge area in the void source. This additionally impacts PD pattern modulation and can be a possible initiation of the formation of cross channels in the bulk isolation which, in the long term, can lead to complete discharge.

Power electronics switching devices currently represent the dominant technology for supplying low voltage (LV) electric motors. The fast switching processes exert a different class of stress on dielectric insulating materials than standard sinusoidal excitations. Such stresses result in an increase in the dynamic activity of the working electric field, which in turn lead to an increased likelihood of partial discharges (PD). The stator design of low voltage motor is often in form of random-wound windings, where the magnet wires (copper or aluminum round wires coated with thin layer of insulation) form a common system of coils with not precisely defined mutual position of particular turns, resulting in various turn-to-turn and coil-to-coil voltage distributions. Pulse Width Modulated (PWM) voltage waveforms from modern electronic inverters are characterized by very short rise times and presence of repetitively occurring overvoltages that can significantly stress the insulation of feeding cables and motors. These factors influence the inception and dynamics of PD and processes of space charge accumulation in electrical insulation. In this paper investigations performed on round magnet wire twisted-pair samples representing LV motor random-wound winding elements are presented. Special attention was afforded to the twist configurations, observed breakdown voltage and PD activity. To describe the field conditions for the formation of PD in the turn-to-turn insulation system, the results of numerical simulations of electric field distributions for winding wires with different diameters, modeled using the COMSOL program, were analyzed. PD created in the insulating systems of model twisted-pair systems were registered and analyzed using the phase resolved partial discharge analysis (PRPDA) method.

This paper describes the influence of AC harmonics in HVDC systems on one of the main indicators of high voltage insulation deterioration, partial discharge. HVDC systems may be subject to various disturbances including harmonics and ripples in the form of pulsating voltages. Experiments, performed on model void specimen at DC voltage with selected superimposed harmonics, are described. A novel approach for recording partial discharges at DC, based on synchronization with selected superimposed harmonics, is proposed. In this way individual partial discharge pulses, occurring during a given measurement time, are accumulated with respect to the phase angles of certain harmonics. The partial discharge inception threshold is analyzed for cases where the DC level is held constant and the AC magnitude and harmonic frequency are varied and for cases where the DC level is varied and the AC magnitude is held constant. In addition, a long term aging test where insulating paper and resin impregnated paper was subjected to pure DC voltage and DC containing 12th harmonic 600 Hz, is described. The aging test shows that the superimposed harmonic content resulted in an amplification of partial discharge presence. By mapping the discharge spots occurring on the void surface, it is possible to quantify the level of erosion in both cases. It was observed that in case of harmonic crest voltage being below partial discharge PD inception level the DC component is crucial, however when harmonics magnitude is above the inception voltage the role of AC is dominant relative to DC. Thus, the presence of harmonics may result in a strong acceleration of the deterioration of insulating materials subjected to DC voltage.

In the resilient and reliable electrical power system, the condition of high voltage insulation plays a crucial role. In the field of high voltage insulation integrity, the partial discharge (PD) inception and development trends are essential for assessment criteria in diagnostics systems. The observed trend to employ more and more sophisticated algorithms with machine learning features and artificial intelligence (AI) elements is observed everywhere. The classification and identification of features in PD images is perceived as a critical requirement for an effective high voltage insulation diagnosis. In this context, techniques allowing for anomaly detection, trends observation, and feature extraction in partial discharge patterns are important. In this paper, the application of few algorithms belonging to image processing, machine learning and optical flow is presented. The feature extraction refers to image segmentation and detection of coherent forms in the images. The anomaly detection algorithms can trigger early detection of the trend changes or the appearance of a new discharge form, and hence are suitable for PD monitoring applications. Anomaly detection can also handle transients and disturbances that appear in the PD image as an indication of an abnormal state. The future monitoring systems should be equipped with trend evolution algorithms. In this context, two examples of insulation aging and application of PD-based monitoring are shown. The first one refers to deep convolutional neural networks used for classification of deterioration stages in high voltage insulation. The latter one demonstrates application of optical flow approach for motion detection in partial discharge images. The motivation for the research was the strive to machine-controlled pattern analysis, leading towards intelligent PD-based diagnostics.