<![CDATA[This paper presents the combination method of wavelet transform and naive bayes classifier to detect and classify the high impedance fault of transmision line. The fault’s current signal is transformed using wavelet. The transformed signal produces coefficients with certain pattern according to the type of fault that occurs. Then, coefficients of transformed signal arevariated become 7 variabels, based on the algorithm of classification. Those variabels are classified using naive bayes classifier to detect and classify the fault of transmission line. Three types of mother wavelet used in this study are Daubechies- 5 (Db5), Daubechies-8 (Db8), and Coiflet-5 (Coif5). Every mother wavelet produces different coefficients. However, they have similar pattern to the algorithm of classification. The highest accuracy of classification was obtained using coeffisients of Daubechies-5 (Db5) at 5th level. The classification accuracy is 97.09% using normal distribution, and 99.78% using kernel distribution.]]>
Copyrights © 2014
