Optimization of SVM Classifier by k-NN for the Smart Diagnosis of the Short-Circuit and Impedance Faults in a PV Generator
Résumé
This paper deals with a new algorithm allowing short-circuit and impedance faults smart diagnosis of PV generators. It is based on the use of the SVM technique for the classification of observations not located in its margin, otherwise the proposed algorithm is used a k-NN method. A PV generator database containing observations distributed over classes is used for testing the new algorithm performance, which shows therefore its contribution and its effectiveness in the diagnosis area. * Parameter j of new observation x *. I' Identity matrix. J Tuning parameter for error accepted. I Current. V Voltage. P Power. PH Photocurrent. I/V Cell Current / Voltage of PV cell. I/V Group Current / Voltage of PV group. I/V Module Current / Voltage of PV module. I/V String Current / Voltage of PV string. I Bypass_Diode Bypass diode current. R s series resistance. t Temperature.
Domaines
Energie électrique
Origine : Fichiers produits par l'(les) auteur(s)
Loading...