TY  - JOUR
T1  - Nodes Variation in Hidden Layer of Partial Discharge Classification
AU - Yusoff, Nur Afifah AU - Isa, Muzamir AU - Adzman, Mohd. Rafi AU - Hafizi Rohani, Mohamad Nur Khairul AU - Yii, Chai Chang AU - Nadiah Ayop, Nurul 
JO  - Research Journal of Applied Sciences
VL  - 11
IS  - 6
SP  - 369
EP  - 373
PY  - 2016
DA  - 2001/08/19
SN  - 1815-932x
DO  - rjasci.2016.369.373
UR  - https://makhillpublications.co/view-article.php?doi=rjasci.2016.369.373
KW  - Partial discharge
KW  -artificial neural network
KW  -classification
KW  -multilayer perceptron
KW  -layer
AB  - This study presents the classification of Partial Discharge (PD) signal using Artificial Neural Network (ANN). This study used the straight forward procedure in PD classification. PD activity is a common element of degradation of the insulation system. The ability of ANN to learn from the example and recognize the pattern between the input data and target output make it interesting tools. Aim of this study discriminates between PD signal and noise signal. The database resulted from laboratory measurement which contains of 3999 data; 2000 data of PD signal and 1999 data of noise signal. Multilayer perceptron with back propagation algorithm is used to develop the network. The nodes of hidden layer is varied from 5-100 with the increment of 5. The result shows that the number of nodes in hidden layer will affect the accuracy of classification. The result of this paper is the average of the 10 combinations produced by the ANN.The classification accuracy of this research is 96.22% for training, 96.10% during testing and mean square error of 0.03039 with the nodes in the hidden layer is 55.
ER  -