S. Borucki

Application possibilities of artificial neural networks for recognizing partial discharges measured by the acoustic emission method

The genesis of the research work presented in this paper constitutes the issue of the effective and efficient recognition of single-source one-time partial discharge forms that can occur in insulation systems of power transformers. The paper presents research results referring to the use of single-direction artificial neural networks for recognizing basic partial discharge forms that can occur in paper-oil insulation impaired by aging processes. The research work results presented show the recognition effectiveness of basic partial discharge forms depending on the descriptor of the analysis of the acoustic emission signal analysis. The detailed cognitive aim was selection of input parameters and an artificial neural network which would be the best, considering recognition effectiveness and processing time, and which could be used as a classifier in an expert diagnostic system making identification of partial discharges measured by using the acoustic method possible.

Diagnosis of Technical Condition of Power Transformers Based on the Analysis of Vibroacoustic Signals Measured in Transient Operating Conditions

The issue of an effective assessment of the technical condition of the mechanical structure of the power transformer core and windings originated the research work discussed in this paper. This publication presents the results which indicate the possibility of developing a new and noninvasive diagnostic method of the transformer active part. The method developed is based on the time-frequency analysis results of the mechanical vibrations measured of the monitored appliance. The registration of vibrations is performed during switching the transformer monitored into idle operation. This paper characterizes the power object under study, the measuring system applied, and discusses a new methodology of assessment of the technical condition of the core and windings using a modified vibroacoustic method. The original measurement results presented in this paper were obtained during switching on a 200-kVA transformer in laboratory conditions. The analysis of the signals measured was carried out for four operating conditions of the transformer under study: 1) with the core pressed by the manufacturer; 2) with the core with loose screws fixing the upper yoke; 3) with the core with loose screws fixing the upper and lower yokes; and 4) with the core with loose screws fixing the upper and lower yokes with a simultaneous unpressing of the windings of the upper voltage side.

Diagnostics of power transformers cores using a modified vibroacoustic method

The paper presents research results referring to the development of a non-invasive method of assessment of the power transformer core technical condition based on the analysis of the registered mechanical vibrations. It characterizes the power unit under study, the measuring system used and the developed methodology of assessment of the core pressing degree using the vibroacoustic method. The original results of the time-frequency analysis of the vibroacoustic signals presented in the paper were obtained during switching on a real 630 kVA dry type power transformer under laboratory conditions. The analysis of the signals registered was carried out for three states of its operation: the core pressed by the manufacturer, the core with loose screws fixing the upper yoke and the core with separated upper yoke beams. The method proposed by authors, the so-called modified vibroacoustic method, is based on the mechanical vibrations measurements of an appliance under study in its transient state - while switching on the unit. The research work aims at developing a modern and, what is most important, non-invasive assessment method of the core technical conditions of power transformers operating in the national electric power system.

Diagnostics of technical condition on load tap changers by acoustic emission method using different types of measuring transducers

Results of scientific research works considering selection of measuring transducer applied for registration of acoustic emission (AE) signals generated by on load tap changers (OLTC) are presented in the paper. During the coupling process AE signals are generated by: the mechanical setup of tap changer, contacts operation and partial discharges. For most OLTC types, the working medium is insulation oil in which the acoustic wave being formed propagates to the steal tank. There exists a possibility to register AE signals with a piezoelectric transducer attached to the outer tank surface or with a hydrophone immersed in the oil. AE signal obtained in such manner contains information which describes operation of the power tap changer and the selector. A comparative analysis of AE signals generated by OLTC working under laboratory conditions in three setups, each applying measuring transducers with different transmittance, is presented in the paper. In measurements three types of transducers were applied: wide band transducer type WD AH-17 from Physical Acoustics Corporation, hydrophone type TC4038 from Reson and accelerometer type 4514 - B - 001 from Brüel & Kjær. The comparative analysis was performed in order to determine their suitability for OLTC technical condition diagnosis. Based on the results achieved one transducer which allows the registration of acoustic events generated by the mechanical setup as well as by electric phenomena occurring during OLTC operation was selected.

High voltage power transformer diagnostics using vibroacoustic method

Processing and analysis results of vibroacoustic signals generated during the switching process of electric power transformer are presented in the paper. Three cases are considered: (1) the transformer has no defects, (2) the pressing screws of the upper yoke were slackened and (3) the pressing screws of both the upper and lower yoke beams were slackened. Further the measuring transducer was placed at four different locations on the transformer tank surface. The gathered signals have been analyzed in the frequency domain by use of the power spectral density function. The magnitude of the first eight harmonics, included in the measured signals, is subjected for comparative analysis.

Using Clustering Methods for the Identification of Acoustic Emission Signals Generated by the Selected Form of Partial Discharge in Oil-Paper Insulation

The article presents the results concerning the use of clustering methods to identify signals of acoustic emission (AE) generated by partial discharge (PD) in oil-paper insulation. The conducted testing featured qualitative analysis of the following clustering methods: single linkage, complete linkage, average linkage, centroid linkage and Ward linkage. The purpose of the analysis was to search the tested series of AE signal measurements, deriving from three various PD forms, for elements of grouping (clusters), which are most similar to one another and maximally different than in other groups in terms of a specific feature or adopted criteria. Then, the conducted clustering was used as a basis for attempting to assess the effectiveness of identification of particular PD forms that modelled exemplary defects of the power transformer’s oil-paper insulation system. The relevant analyses and simulations were conducted using the Matlab estimation environment and the clustering procedures available in it. The conducted tests featured analyses of the results of the series of measurements of acoustic emissions generated by the basic PD forms, which were obtained in laboratory conditions using spark gap systems that modelled the defects of the power transformer’s oil-paper insulation.

Application of multi-comparational algorithms to classification of acoustic emission signals generated by partial discharges

This article concerned on the mechanism of functioning of two multi-comparational algorithms, that is: CMCA (Classic Multi-Comparational Algorithm) and OMCA (Optimised Multi-Comparational Algorithm). The range and possibility of application of frequency descriptors that explicitly characterise the acoustic emission (AE) signals from partial discharges (PD) were described. The method for selection of quantity of the standard waveforms for the purpose of the classification process, carried out using the tested algorithms was presented. In addition, there were showed results of the simulation, performed using the proposed algorithms and two databases containing the AE signal waveforms generated by basic forms of PD, which can occur in insulation systems of transformers. The efficiency of CMCA and OMCA algorithms in application for AE signals generated by PD, as well as the reason to complete the developed database by another class of waveforms due to the defect of insulation system that not being catalogued previously was demonstrated in the study.

The analysis of the impact point of the power transformer core of torsional load on the measured parameters of the vibroacoustics signals

The research, whose results are presented in this paper, was conducted in the Institute of Electric Power Engineering of Opole University of Technology and focused on the development of the vibroacoustic method used to evaluate the condition of the mechanical structure of the magnetic circuit units operating in the national transformer system. The paper presents results of research work related to the development of the vibroacoustic method to evaluate the technical condition of power transformer cores. The paper presents the results of laboratory measurements and analysis of mechanical vibrations in the transformer core according to the variable torque twisting the screw yoke of the monitored unit. The aim of this work was to evaluate the torque twisting the yoke of the diagnosed object on the registered parameters of vibroacoustic signals which are used in the diagnosis of power transformers, magnetic circuits.

Parameters of acoustic emission signals constituting the knowledge base of a computer diagnostic system of power transformer insulation systems

The paper presents the range and possibilities of using expert systems at diagnostic measurements of insulation systems of power appliances taken by the acoustic emission (AE) method. Multi-parametric descriptors characterizing the AE signals, which can be used in the knowledge base as criteria making the identification of basic partial discharge forms (PDs) possible, have been characterized in detail. The paper shows descriptors describing the AE signals registered in time, frequency, time-frequency domains and statistical data.