2008 · 13 citations · 11 references
ComputingComputer ScienceElectrical EngineeringData MiningPower ElectronicsTime Series AnalysisFault DetectionKernel MethodsPower Grid AnalysisSmart Grid SystemsSmart GridsSupport Vector MachinesFault ClassificationFault Detection Using Machine LearningFault LocationMachine Learning For Fault DetectionMachine Learning-based Fault DiagnosisSimulink ModelsSupport Vector ClassificationSupport Vector Machine Classification
This paper presents a practical approach to identify the location of fault in a distribution system. The proposed approach uses measurements available at the substations. Whenever a fault occurs in a system, the direction as well as the magnitude of power change in each phase from the normal condition is used as the input data for the algorithm. Based on this information, support vector machine (SVM) locates fault location that is emanating from a substation end. After identifying the fault location, the events occurred in the phases is then classified individually with the help of two-stage fuzzy classifier. The results obtained on a practical distribution (radial) system simulated in Matlab simulink and PSCAD show the effectiveness of the proposed algorithm.
11
Yuhai Wu, Vladimir Vapnik · Technometrics · 1999 · 27K citations
Electrical power systems quality
M. McGranaghan, R.C. Dugan, H. Wayne Beaty · Choice Reviews Online · 1996 · 2.6K citations · Full text