Xue-min Zhang
School of Computer and Information Science, Hubei Engineering University, Xiaogan City, 432000, China
Zeng-gang Xiong
School of Computer and Information Science, Hubei Engineering University, Xiaogan City, 432000, China
Yao-ming Ding
School of Computer and Information Science, Hubei Engineering University, Xiaogan City, 432000, China
Qin-qin Zhang
Machine Learning and Cybernetics Research Center, School of Computer Science and Engineering, South China University of Technology, Guangzhou 510006, China
Jing Zhu
Machine Learning and Cybernetics Research Center, School of Computer Science and Engineering, South China University of Technology, Guangzhou 510006, China
ABSTRACT
Transmission line is an important part of smart grid, so it very significant to accurately and quickly detect the occurrence of the fault location once the fault occurs in transmission line. Currently, there are many artificial intelligence methods for fault location, but most of them use a single classifier, such as SVM, RBFNN, BPNN and so on. Many scholars have demonstrated Multiple Classifier Systems (MCSs) are more accurate than the single classification system in many application areas. In this study, we will propose a classifier selection method which depends on member classifiers relative error rate and sensitivity. In this method, we will try to find the nearest the training sample of the unseen sample, then attempt to select a most suitable member classifier for the testing sample according to the base classifiers testing error for the nearest training sample and the sensitivity of the classifiers for the unseen sample and the training sample. At last, we will use the selected classifier predict the unseen sample. In this paper, five kinds of smart grid fault location problem data set will be used for our experiment and we will illustrate the comparative of our proposed classifier selection method.
PDF References Citation
How to cite this article
Xue-min Zhang, Zeng-gang Xiong, Yao-ming Ding, Qin-qin Zhang and Jing Zhu, 2013. A Novel Dynamic Classifier Selection for Transmission Line Fault Location. Journal of Applied Sciences, 13: 2051-2056.
DOI: 10.3923/jas.2013.2051.2056
URL: https://scialert.net/abstract/?doi=jas.2013.2051.2056
DOI: 10.3923/jas.2013.2051.2056
URL: https://scialert.net/abstract/?doi=jas.2013.2051.2056
REFERENCES
- Jiang, J.A., C.L. Chuang, Y.C. Wang, C.H. Hung, J.Y. Wang, C.H. Lee and Y.T. Hsiao, 2011. A hybrid framework for fault detection, classification and location-Part I: Concept, structure and methodology. IEEE Trans. Power Delivery, 26: 1988-1998.
CrossRef - Shahid, N., S.A. Aleem and I.H. Naqvi, 2012. Support vector machine based fault detection & classification in smart grids. Proceedings of the Globecom Workshops, December 3-6, 2012, Anaheim, CA., pp: 1526-1531.
CrossRef - Silva, K.M., B.A. Souza and N.S.D. Brito, 2006. Fault detection and classification in transmission lines based on wavelet transform and ANN. IEEE Trans. Power Delivery, 21: 2058-2063.
CrossRef - Samantaray, S.R., P.K. Dash and G. Panda, 2006. Fault classification and location using HS-transform and radial basis function neural network. Elect. Power Syst. Res., 76: 897-905.
CrossRefDirect Link - Samantaray, S.R., P.K. Dash and G. Panda, 2007. Distance relaying for transmission line using support vector machine and radial basis function neural network. Elect. Power Energy Syst., 29: 551-556.
CrossRefDirect Link - Yeung, D.S. and P.P.K. Chan, 2009. A novel dynamic fusion method using localized generalization error model. Proceedings of the International Conference on Systems, Man and Cybernetics , October 11-14, 2009, San Antonio, pp: 623-628.
CrossRef - Woods, K., W.P. Kegelmeyer Jr. and K. Bowyer, 1997. Combination of multiple classifiers using local accuracy estimates. IEEE Trans. Patt. Anal. Mach. Intelli., 19: 405-410.
Direct Link - Kittler, J., M. Hatef, R.P.W. Duin and J. Matas, 1998. On combining classifiers. IEEE Trans. Pattern Anal. Mach. Intell., 20: 226-239.
CrossRef - Yeung, D.S., W.W.Y. Ng and D. Wang, 2007. Localized generalization error model and its application to architecture selection for radial basis function neural network. IEEE Trans. Neural Network, 18: 1294-1305.
CrossRef