The Transactions of The Korean Institute of Electrical Engineers (전기학회논문지)

The Korean Institute of Electrical Engineers (대한전기학회)
권호 표지
DOI QR코드

DOI QR Code

Design of Partial Discharge Pattern Classifier of Softmax Neural Networks Based on K-means Clustering : Comparative Studies and Analysis of Classifier Architecture

K-means 클러스터링 기반 소프트맥스 신경회로망 부분방전 패턴분류의 설계 : 분류기 구조의 비교연구 및 해석

  • Jeong, Byeong-Jin (School of Electrical & Electronic Engineering, the University of Suwon) ;
  • Oh, Sung-Kwun (School of Electrical & Electronic Engineering, the University of Suwon)
  • Received : 2017.11.13
  • Accepted : 2017.11.30
  • Published : 2018.01.01
Download PDF
⟨ Previous Next ⟩

Abstract

This paper concerns a design and learning method of softmax function neural networks based on K-means clustering. The partial discharge data Information is preliminarily processed through simulation using an Epoxy Mica Coupling sensor and an internal Phase Resolved Partial Discharge Analysis algorithm. The obtained information is processed according to the characteristics of the pattern using a Motor Insulation Monitoring System program. At this time, the processed data are total 4 types that void discharge, corona discharge, surface discharge and slot discharge. The partial discharge data with high dimensional input variables are secondarily processed by principal component analysis method and reduced with keeping the characteristics of pattern as low dimensional input variables. And therefore, the pattern classifier processing speed exhibits improved effects. In addition, in the process of extracting the partial discharge data through the MIMS program, the magnitude of amplitude is divided into the maximum value and the average value, and two pattern characteristics are set and compared and analyzed. In the first half of the proposed partial discharge pattern classifier, the input and hidden layers are classified by using the K-means clustering method and the output of the hidden layer is obtained. In the latter part, the cross entropy error function is used for parameter learning between the hidden layer and the output layer. The final output layer is output as a normalized probability value between 0 and 1 using the softmax function. The advantage of using the softmax function is that it allows access and application of multiple class problems and stochastic interpretation. First of all, there is an advantage that one output value affects the remaining output value and its accompanying learning is accelerated. Also, to solve the overfitting problem, L2-normalization is applied. To prove the superiority of the proposed pattern classifier, we compare and analyze the classification rate with conventional radial basis function neural networks.

Keywords

References

  1. S. M. Hassan Hosseini, S. M. Hosseini Bafghi, "Comparison of High Frequency Detailed Generator Models for Partial Discharge Localization", Journal of Electrical Engineering & Technology 10, 1752-1758, 2015.7. https://doi.org/10.5370/JEET.2015.10.4.1752
  2. Lianwei Bao, Jian Li, Jing Zhang, Tianyan Jiang, "Partial Discharge Process and Characteristics of Oil-Paper Insulation under Pulsating DC Voltage", Journal of Electrical Engineering & Technology 11, 437-445, 2016. 03.
  3. Furqan Asghar, Muhammad Talha, Se-Yoon Kim, SungHo Kim, "Hotelling T2 Index Based PCA Method for Fault Detection in Transient State Processes", Journal of Institute of Control, Robotics and Systems 22, 276-280, 2016.4. https://doi.org/10.5302/J.ICROS.2016.16.8006
  4. Sung-Kwun Oh, Wook-Dong Kim, Witold Pedrycz, Su-Chong Joo, "Design of K-means clustering-based polynomial radial basis function neural networks (pRBF NNs) realized with the aid of particle swarm optimization and differential evolution", Neurocomputing, Vol. 78, No. 1, pp. 121-132, 2012. https://doi.org/10.1016/j.neucom.2011.06.031
  5. Mingwen Zheng, Yanping Zhang, "An Algorithm to Determine RBFNN's Center Based on the Improved Density Method", Open Journal of Applied Sciences, 1-5, 2014. 04
  6. John E. Dennis, David M. Gay, Roy E. Walch, "An Adaptive Nonlinear Least-Squares Algorithm", ACM Transactions on Mathematical Software, Vol. 7, No. 3, pp. 348-368. 1981. https://doi.org/10.1145/355958.355965
  7. Xiong Luo, Xiao hui Chang, Xiaojuan Ban "Regression and classification using extreme learning machine based on L1-norm and L2-norm", Neurocomputing, Vol 174, Part A, pp. 179-186, 2016. https://doi.org/10.1016/j.neucom.2015.03.112
  8. Shenglei Chen, Geng Chen, Ruijun Gu, "An efficient L2-norm regularized least-squares temporal difference learning algorithm", Knowledge-Based Systems, Vol 45, pp. 94-99, 2013. https://doi.org/10.1016/j.knosys.2013.02.010
  9. Kim, Joong-Hyo, Kwon, Sung-Dae, Hong, Jeong-Pyo, Ha, Tae-Jun, "Development of Optimum Traffic Safety Evaluation Model Using the Back-Propagation Algorithm", JOURNAL OF THE KOREAN SOCIETY OF CIVIL ENGINEERS 35(4), 2015.6, 679-690. https://doi.org/10.12652/Ksce.2015.35.3.0679
  10. Jinju Hong, Sehan Kim, Jeawon Park, Jaehyun Choi, "A Malicious Comments Detection Technique on the Internet using Sentiment Analysis and SVM", Journal of the Korea Institute of Information and Communication Engineering 20, 260-267, 2016.2. https://doi.org/10.6109/jkiice.2016.20.2.260
  11. Byeong-Jin Jeong, Seung-Cheo Leel, Sung-Kwun Oh, "Design of Optimized Radial Basis Function Neural Networks Classifier Using EMC Sensor for Partial Discharge Pattern Recognition", The Korean Institute of Electrical Engineers 66(9), 1392-1401, 2017.9.