TY - GEN
T1 - Comparison of Machine Learning Algorithms for Classification of Partial Discharge Signals in Medium Voltage Components
AU - Kumar, Haresh
AU - Shafiq, Muhammad
AU - Hussain, Ghulam Amjad
AU - Kauhaniemi, Kimmo
N1 - Funding Information:
This work was carried out first with the support of the Nissi Foundation scholarship, and later it was supported by the Fortum grant. The second author acknowledges the Estonian Research Council's grant under project No. PSG632.
Publisher Copyright:
© 2021 IEEE.
PY - 2021
Y1 - 2021
N2 - Partial discharge (PD) diagnosis is an effective tool to track the condition of electrical insulation in the medium voltage (MV) power components. Machine Learning Algorithms (MLAs) promote automated diagnosis solutions for large scale and reliable maintenance strategy. This paper aims to investigate the performance of two MLAs: Support Vector Machine (SVM) and K-Nearest Neighbors (KNN) for the classification of different types of PD sources. Suitable features are extracted by applying statistical parameters on the coefficients of discrete wavelet transform (DWT) for observing the performance of both MLAs. The performance of the algorithms is evaluated using key performance indicators (KPIs); accuracy, prediction speed and training time. Besides KPIs, a confusion matrix is presented to highlight the accurately classified and misclassified PD signals for the SVM algorithm. Comparative study of both algorithms demonstrates that SVM provides better results as compared to the KNN algorithm. The proposed solution can be valuable for the development of automated classification.
AB - Partial discharge (PD) diagnosis is an effective tool to track the condition of electrical insulation in the medium voltage (MV) power components. Machine Learning Algorithms (MLAs) promote automated diagnosis solutions for large scale and reliable maintenance strategy. This paper aims to investigate the performance of two MLAs: Support Vector Machine (SVM) and K-Nearest Neighbors (KNN) for the classification of different types of PD sources. Suitable features are extracted by applying statistical parameters on the coefficients of discrete wavelet transform (DWT) for observing the performance of both MLAs. The performance of the algorithms is evaluated using key performance indicators (KPIs); accuracy, prediction speed and training time. Besides KPIs, a confusion matrix is presented to highlight the accurately classified and misclassified PD signals for the SVM algorithm. Comparative study of both algorithms demonstrates that SVM provides better results as compared to the KNN algorithm. The proposed solution can be valuable for the development of automated classification.
KW - classification
KW - electrical insulation
KW - features extraction
KW - key performance indicators
KW - machine learning algorithms
KW - partial discharge
UR - http://www.scopus.com/inward/record.url?scp=85123908491&partnerID=8YFLogxK
U2 - 10.1109/ISGTEurope52324.2021.9639923
DO - 10.1109/ISGTEurope52324.2021.9639923
M3 - Conference contribution
T3 - Proceedings of 2021 IEEE PES Innovative Smart Grid Technologies Europe: Smart Grids: Toward a Carbon-Free Future, ISGT Europe 2021
BT - Proceedings of 2021 IEEE PES Innovative Smart Grid Technologies Europe
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 11th IEEE PES Innovative Smart Grid Technologies Europe, ISGT Europe 2021
Y2 - 18 October 2021 through 21 October 2021
ER -
