Title :
Fault Identification in Induction Motors with RBF Neural Network Based on Dynamical PCA
Author :
Kilic, Erdal ; Ozgonenel, Okan ; Ozdemir, Ali Ekber
Author_Institution :
Ondokuz Mayis Univ., Samsun
Abstract :
Early detection and diagnosis of incipient faults is desirable for on-line condition assessment, product quality assurance and improved efficiency of induction motors running off power supply mains. In the applications of three-phase induction motors in industry, the inner faults may occur in their rotor and stator windings. These kinds of faults will make serious health problems on the motor. This paper presents a new protection scheme for internal short circuit faults occurring with a degree (single or multiple) in three-phase induction motors. The results are compared with traditional outcomes existed from fast Fourier transformation (FFT) of the motor currents. The proposed algorithm is simpler and only uses stator currents. There is no need any other sensor knowledge.
Keywords :
electric machine analysis computing; fault diagnosis; induction motor protection; principal component analysis; radial basis function networks; short-circuit currents; RBF neural network; dynamical PCA; fault identification; induction motors; on-line condition assessment; power supply mains; product quality assurance; protection scheme; short circuit faults; Circuit faults; Electricity supply industry; Fault detection; Fault diagnosis; Induction motors; Neural networks; Power supplies; Principal component analysis; Quality assurance; Rotors; FFT; Internal faults; fault identification (FI); induction motor; principal component analysis (PCA); radial basis functions (RBFs);
Conference_Titel :
Electric Machines & Drives Conference, 2007. IEMDC '07. IEEE International
Conference_Location :
Antalya
Print_ISBN :
1-4244-0742-7
Electronic_ISBN :
1-4244-0743-5
DOI :
10.1109/IEMDC.2007.382776
