Application des fonctions Kernels de la méthode LS-SVM pour le diagnostic d’un isolateur HT pollué
This work presents a method to predict the polluted level of the surfaces of an insulator, that is to say, to diagnose the operational conditions of the isolation of an electrical system by pattern recognition techniques using some types of methods such as Least square support vectors machines (LS-SVM); we present here several kernel functions like RBF, polykernel and MLP. The methodology is to use as input variables of the insulation such as diameter, height, creepage line, form factor and equivalent salt deposition density. The majority of the variables to be predicted are dependent on several independent variables. The results of this work are useful in predicting the severity of contamination, the critical overvoltage; arc length and especially affects the overvoltage. The validity of the approach was examined by testing several insulators with different geometries. Field experience and laboratory tests are expensive both in time and money; therefore this method takes efficiency vs experimental tests in laboratories. A comparison of the kernel functions used shows the improvement of LS-SVM with RBF, Polykernels and that the use of combined models is a powerful technique for this type of application demand.
B. Zegnini, A.H. Mahdjoubi, M. Belkheiri, A least squares support vectors machines (LS-SVM) approach for predicting critical flashover voltage of polluted insulators, CEIDP , 403-406, 2011.
M. Belkheiri, B. Zegnini and D. Mahi, Modeling of the critical flashover voltage of high voltage insulators using artificial intelligence, Journal of intelligent Computing and Applications, Serial publications 2(2) 137-154,2009.
M. T. Gencoglu , M. Uyar , Prediction of flashover voltage of insulators using least squares support vector machines, Expert Systems with Applications 36 10789–10798, 2009.
V.T. Kontagyri, A.A. Gialketsi, G.J. Tsekouras, I.F. Gonos, I.A. Stathopulos, Design of an artificial neural network for the estimation of the flashover voltage on insulators, Electric power systems research 77, 1532-1540, 2007.
V. N. Vapnik, Statistical learning theory, Wiley 1998.
J. A. K Suykens, T V. Gestel, J. D. Brabanter, B. De Moor, J. Vandewalle, Least Squares support vector machines. Singapore: World Scientific, 2002.
M.T.Gencoglu, M. Cebeci, The pollution flashover on high voltage insulators, Electr. Power Syst. Res. 78 (11), 1914–1921, 2008.
Abdelhalim Mahdjoubi, Tahar Chenaf, Boubakeur Zegnini, Mohammed Belkheiri, «Power Transformer Fault Diagnosis Based On Support Vector Machine (Svm) », Cagre’19, Algiers, Algeria, February 26-28, 2019
A. Mahdjoubi, « Utilisation De L’approche Des Machines A Vecteurs De Support (Svm) Pour La Prédiction Des Performances Electriques Des Systemes D’isolation Utilisés Dans Les Réseaux De Transport D’énergie » Thèse De Doctorat Soutenu Septembre 2014, UATL
Mahdjoubi A, Zegnini B, Belkheir M. Prediction Of Critical Flashover Voltage Of Polluted Insulators Under Sec And Rain Conditions Using Least Squares Support Vector Machines (Ls-Svm). Diagnostyka. 2019;20(1):49-54. Https://Doi.Org/10.29354/Diag/99854
Ah. Mahdjoubi, B. Zegnini, M. Belkheiri, «A New Approach Using Least Squares Support Vector Machines (Ls-Svm) To Predict Furan In Power Transformers », Przegląd Elektrotechniczny, Pologne, 2014, Pp: 142-145.