Title :
Predicting behavior of induction motors during service faults and interruptions
Author_Institution :
Power Electron. Applications Centre, Knoxville, TN, USA
Abstract :
A neural network-based identification for induction motor speed is proposed. The backpropagation neural network technique is used to provide real-time adaptive estimation of the motor speed. The validity and effectiveness of the proposed estimator as well as its sensitivity to parameter variation are verified by digital simulations. The proposed identification performs well under vector control and therefore can lead to an improvement in the performance of speed sensorless drives. The new approach is presented in a way that will contribute to a better understanding of neural network applications to motion control
Keywords :
adaptive control; backpropagation; digital control; digital simulation; electric machine analysis computing; induction motor drives; machine control; motion control; neurocontrollers; parameter estimation; real-time systems; velocity control; applications; backpropagation; behavoiur prediction; digital simulations; induction motors; interruptions; motion control; neural network; parameter variation; performance; real-time adaptive speed estimation; sensitivity; service faults; vector control; Application software; Induction motors; Magnetic devices; Military computing; Motor drives; Power conditioning; Power system protection; Process control; Production facilities; Voltage fluctuations;
Journal_Title :
Industry Applications Magazine, IEEE
DOI :
10.1109/2943.378057
