IEEE Transactions on Dielectrics and Electrical Insulation · 2019 · 10 citations · 21 references
Natural SciencesElectrical EngineeringMaintenance EngineeringPower ElectronicsStatistical ModelingDependence ModelingJoint DistributionsMultivariate AnalysisMultivariate StatisticsPower QualityProbability DistributionsBivariate DistributionsBreakdown VoltageCopula ModelsCopula-based Dependence ModelingDependence StructureGoodness-of-fitGoodness-of-fit AssessmentHigh-voltage Breakdown
This paper aims at investigating the applicability of copula approach to accommodate the dependence structure between total combustible gases (TCG) and breakdown voltage (BDV). The TCG and BDV have been obtained from dissolved gas investigation and oil parameter testing respectively by utilizing the real-time data of transformer and reactors in service for more than a decade. In the proposed study, considering the negative association and goodness-of-fit measure, the Frank copula has been found satisfactory to model the bivariate joint distribution. Hence it has been used to develop the joint and conditional probability between them. Lastly, the outcome of this novel analysis validates the relationship between TCG and BDV which can be utilized for assessing the influence of total combustible gases on oil parameters. This will further help in taking precautionary measures for maintaining the quality of the oil for retaining the healthiness of transformers and reactors.
21
A NEW MEASURE OF RANK CORRELATION
M. G. Kendall · Biometrika · 1938 · 5.8K citations
A New Measure of Rank Correlation
M. G. Kendall · Biometrika · 1938 · 2.3K citations