New correlation features for dissolved gas analysis based transformer fault diagnosis based on the maximal information coefficient

Liang, Yongliang; Zhang, Zhongyi; Li, Kejun; Li, Yu-Chuan

doi:10.1049/hve2.12136

Cited by 16 publications

(11 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In data statistics, the MIC analysis method was used to analyse the statistical data. The reason behind this choice is that MIC belongs to the largest non-parametric exploration method based on information, and it has the basic characteristics of universality, fairness and symmetry with higher robustness, wider applicability and more accuracy than Pearson and Spearman statistical methods [27,28]. Therefore, this method is more suitable for acquiring the feature of statistical data.…”

Section: Statistical Datamentioning

confidence: 99%

Dispersion characteristics of switching transients from large‐scale experiments in a 1100 kV gas insulated switchgear circuit

et al. 2023

View full text Add to dashboard Cite

The switching operation of gas insulated switchgear (GIS) disconnector will produce multiple random spark discharges between the gas gaps of the disconnector. Each spark discharge can be affected by various random factors, such as trapped charge on the load side, the initial operating phase, gap distance and operating speed of the disconnector, leading to dispersion characteristics. At present, some studies have mentioned this dispersion, but its impact is still unclear due to the lack of experimental data, resulting in a deviation between the simulation results and the measured results. In this study, thousands of switching operations were carried out based on the 1100 kV full‐scale GIS circuit in Wuhan ultra‐high voltage alternating current Test Base. The maximal information coefficient analysis, linear regression fitting and other statistical methods were used to discuss the dispersion characteristics of very fast transient overvoltage (VFTO) and transient enclosure voltage (TEV) from large‐scale experimental data. The results show that the dispersion of VFTO and TEV in amplitude aspects both increase gradually as the breakdown voltage increases, with the maximum difference of five times of TEV amplitude at normalised breakdown voltage. In terms of frequency characteristics, the dispersion of different frequency components in TEV is always greater than that of VFTO, especially for the high‐frequency components. Moreover, by comparing the frequency characteristics of TEV at different positions, we notice that the dispersion is almost independent of the spatial position but only determined by the randomness of spark discharges. These discoveries reveal the importance of the dispersion in switching transients and make up for the lack of theoretical understanding of the correlation among them.

show abstract

Section: Statistical Datamentioning

confidence: 99%

Dispersion characteristics of switching transients from large‐scale experiments in a 1100 kV gas insulated switchgear circuit

et al. 2023

View full text Add to dashboard Cite

show abstract

“…Generally, dissolved gas analysis (DGA) is used to analyse gases in transformer oil [16,17]. It is one of the most important and effective techniques for assessing the condition of oil-filled transformers [18].…”

Section: Introductionmentioning

confidence: 99%

Dissolved gas analysis and dissipation factor measurement of mineral oil‐based nanofluids under thermal and electrical faults

et al. 2023

View full text Add to dashboard Cite

Mineral oil is the most frequent insulating liquid used in oil‐immersed transformers for electrical insulation and heat dissipation. However, oil‐based nanofluids are becoming more popular in scientific research as they have proved to have better dielectric and thermal characteristics. When applying these nanofluids into actual transformers, they would be exposed to thermal and electrical stresses. Thus, The aim of the authors is to investigate the generation pattern of dissolved gases in nanofluids under thermal and electrical faults and to assess the applicability of traditional Dissolved Gas Analysis (DGA) techniques if oil‐based nanofluids are used in transformers. Oil‐based nanofluid samples were prepared using a magnetic stirrer and an ultrasonic homogeniser and then subjected to simulated thermal and electrical faults in the laboratory using properly sealed test cells. Three types of metal oxides, Silicon dioxide, Titanium dioxide, and Aluminium oxide nanoparticles, have been used to prepare nanofluids with 0.02 g/L and 0.04 g/L concentrations. The gases released and dissolved into oil due to the simulated faults were analysed and compared to traditional mineral oil as a benchmark. The dielectric dissipation factor was obtained and analysed for all samples. According to the findings, the presence and concentration of nanoparticles were shown to influence the amount of gases produced. As a result, this research is crucial in guaranteeing that traditional DGA techniques can be employed in transformers that use oil‐based nanofluids.

show abstract

“…Among these techniques, DGA is one of the most widely used techniques. Preventive maintenance tool for power transformers, the DGA is used for condition monitoring, early-stage detection of faults in active parts and unplanned outage prevention [6,7].…”

Section: Introductionmentioning

confidence: 99%

Improved intelligent methods for power transformer fault diagnosis based on tree ensemble learning and multiple feature vector analysis

Hechifa,

Lakehal,

Nanfak

et al. 2023

Electr Eng

View full text Add to dashboard Cite

This paper discusses the impact of the feature input vector on the performance of DGA-based intelligent power transformer fault diagnosis methods. For this purpose, 22 feature vectors from traditional diagnostic methods were used as feature input vectors for four tree-based ensemble algorithms, namely random forest (RF), tree ensemble (TE), gradient boosted tree (GBT), and extreme gradient tree (XGB). To build the proposed diagnostics models, 407 samples were used for training and testing. For validation and comparison with the existing methods of literature 89 samples were used. Based on the results obtained on the training and testing datasets, the best performance was achieved with feature vector 16, which consists of the gas ratios of Rogers' four ratios method and the three ratios technique. The test accuracies based on these vectors are 98.37, 96.75, 95.93, and 97.56% for the RF, TE, GBT, and XGB algorithms, respectively. Furthermore, the performance of the methods based on best input feature were evaluated and compared with other methods of literature such as Duval Triangle, modified Rogers' four ratios method, combined technique, three ratios technique, Gouda triangle, IEC 60599, NBR 7274, clustering, and key gases with gas ratio methods. On validating dataset, diagnostic accuracies of 92.13, 91.01, 89.89, and 91.01% were achieved by the RF, TE, GBT, and XGBoost models, respectively. These diagnostic accuracies are higher than 83.15 % of the clustering method and 82.02 % of combined technique which are the best existing methods. Even if the performance of DGA-based intelligent methods depends strongly on the shape of the feature vector used, this study provides scholars with a tool for choosing the feature vector to use when implementing these methods.

show abstract

New correlation features for dissolved gas analysis based transformer fault diagnosis based on the maximal information coefficient

Cited by 16 publications

References 32 publications

Dispersion characteristics of switching transients from large‐scale experiments in a 1100 kV gas insulated switchgear circuit

Dispersion characteristics of switching transients from large‐scale experiments in a 1100 kV gas insulated switchgear circuit

Dissolved gas analysis and dissipation factor measurement of mineral oil‐based nanofluids under thermal and electrical faults

Improved intelligent methods for power transformer fault diagnosis based on tree ensemble learning and multiple feature vector analysis

Contact Info

Product

Resources

About