A Relevance Vector Machine and Bare-bones Particle Swarm Optimization Hybrid Algorithm for PD Pattern Recognition of XLPE Cable

In the paper, a relevance vector machine and bare-bones particle swarm optimization hybrid algorithm (BBPSO-RVM) is rst presented and used to recognize the partial discharge patterns in XLPE power cables, and the bare-bones particle swarm optimization (BBPSO) is proposed to select the parameters of relevance vector machine (RVM). In the study, 200 cases are collected to testify the eectiveness of the relevance vector machine and bare-bones particle swarm optimization hybrid algorithm compared with other classiers. Traditional relevance vector machine classier, support vector machine classier and RBF neural network are used to compare with the relevance vector machine and bare-bones particle swarm optimization hybrid algorithm. Finally, the outputs of the relevance vector machine and bare-bones particle swarm optimization hybrid algorithm, relevance vector machine classier, support vector machine classier (SVM) and RBF neural network (RBFNN) are given, and the comparison of PD pattern recognition accuracies of XLPE cable among BBPSO-RVM, RVM, SVM and RBFNN is given. The experimental results show that recognition accuracy of BBPSO-RVM is higher than that of RVM, SVM and RBFNN. In the study, 200 cases are collected to testify the eectiveness of the relevance vector machine and bare-bones particle swarm optimization hybrid algorithm compared with other classiers.Tradit ional relevance vector machine classier, support vector machine classier and RBF neural network are used to compare with the relevance vector machine and bare-bones particle swarm optimization hybrid algorithm. The outputs of the relevance vector machine and barebones particle swarm optimization hybrid algorithm, relevance vector machine classier, support vector machine classier and RBF neural network are given. It can be seen from the comparison of PD pattern recognition accuracies of XLPE cable among BBPSO-RVM, RVM, SVM and RBFNN that recognition accuracy of BBPSO-RVM is higher than that of RVM, SVM and RBFNN.

[1]  Elias Zintzaras,et al.  Forest classification trees and forest support vector machines algorithms: Demonstration using microarray data , 2010, Comput. Biol. Medicine.

[2]  Leandro dos Santos Coelho,et al.  Gaussian quantum-behaved particle swarm optimization approaches for constrained engineering design problems , 2010, Expert Syst. Appl..

[3]  Euntai Kim,et al.  A new weighted approach to imbalanced data classification problem via support vector machine with quadratic cost function , 2011, Expert Syst. Appl..

[4]  Reza Akbari,et al.  A rank based particle swarm optimization algorithm with dynamic adaptation , 2011, J. Comput. Appl. Math..

[5]  Aris Kornelakis,et al.  Multiobjective Particle Swarm Optimization for the optimal design of photovoltaic grid-connected systems , 2010 .

[6]  Sugata Munshi,et al.  Correlation technique and least square support vector machine combine for frequency domain based ECG beat classification. , 2010, Medical engineering & physics.

[7]  Taskin Kavzoglu,et al.  A kernel functions analysis for support vector machines for land cover classification , 2009, Int. J. Appl. Earth Obs. Geoinformation.

[8]  Cheng-Jian Lin,et al.  Pattern recognition using neural-fuzzy networks based on improved particle swam optimization , 2009, Expert Syst. Appl..

[9]  Sun Choi,et al.  In silico classification of adenosine receptor antagonists using Laplacian-modified naïve Bayesian, support vector machine, and recursive partitioning. , 2010, Journal of molecular graphics & modelling.

[10]  Engin Avci.  Selecting of the optimal feature subset and kernel parameters in digital modulation classification by using hybrid genetic algorithm-support vector machines: HGASVM , 2009, Expert Syst. Appl..

[11]  Hakan Cevikalp,et al.  New clustering algorithms for the support vector machine based hierarchical classification , 2010, Pattern Recognit. Lett..