Zinger, Donald S.
M.S. (Master of Science)
Department of Electrical Engineering
Electric machinery; Induction; Electric machinery--Monitoring; Harmonic analysis; Speed
Sensorless speed detection of PWM-fed induction motor from the harmonic content of the current spectrum is presented in this work. Interleaving the theoretical algorithm research with practical implementation aspects to find efficient and cost effective techniques to extract the speed of an induction motor has been the principal goal of this work. In order to achieve this, a comprehensive study of digital signal processing algorithms and their feasibility to extract the speed-related harmonics in the current spectrum has been done. The main emphasis has been laid on the quest of algorithms which could provide reliable means of speed detection updates at regular intervals even for applications with fluctuating speeds. Special emphasis has been paid for success of the algorithms at low-frequency operations where speed detection becomes a challenging task. To evaluate the speed-related harmonics, current spectrum has been chosen over its voltage counterpart. The eccentricity harmonics and the rotor slot harmonics available in the current spectrum have been studied in depth and their mathematical and physical relationship to the speed of the machine have been studied and presented. The dataacquisition time requirements and the problems associated for a successful implementation using the eccentricity harmonics have been investigated for machines operating in nonsteady-state conditions. Eccentricity harmonics have been found insufficient for good speed detection unless the frequency resolution is improved. This is true especially for applications with fluctuating speeds. An aliasing algorithm has been presented which familiarizes itself with the available rotor slot harmonics in the spectrum. The harmonics identified are enhanced using aliasing techniques. Further, these enhanced harmonics are subjected to spectral estimation techniques which helps to remove any spectral ambiguity if at all present. The spectral estimation has shown to resolve the problem of large data acquisition and also has shown significant improvement for implementations in which the motor does not operate in steady state for long periods. Real-time feasibility has been studied in terms of the data acquisition, economy, signal processing, memory requirements, etc., to give a fast and reliable speed update.
Batra, Ashish, "Sensorless speed detection of PWM-fed asynchronous machines using spectral estimation techniques" (1997). Graduate Research Theses & Dissertations. 4877.
xiii, 145 pages
Northern Illinois University
Rights Statement 2
NIU theses are protected by copyright. They may be viewed from Huskie Commons for any purpose, but reproduction or distribution in any format is prohibited without the written permission of the authors.