Skip to Main Content
 

Global Search Box

 
 
 

ETD Abstract Container

Abstract Header

Detection and Pattern Recognition of Partial Discharge in Electric Machine Coils with Pulsed Voltage Excitation

Xiong, Han
http://rave.ohiolink.edu/etdc/view?acc_num=osu1555592971342072

Abstract Details

2019, Doctor of Philosophy, Ohio State University, Electrical and Computer Engineering.
For over 80 years, partial discharges (PDs) have been monitored in traditional 50-/60-Hz medium- to high-voltage electric machines to detect and determine various types of defects in the insulation system. The PD pattern and changing trend revealed by the continuous monitoring have also been used as an indicator of insulation health condition. With the development of modern power electronics, an increasing amount of medium- to high-voltage electric machines are now driven by variable speed drives (VSDs) for higher efficiency and better control capability. The voltage excitation, instead of the traditional low-frequency sinusoidal voltage, becomes high-frequency pulsed voltages, or pulse-width-modulated (PWM) voltages. The short rise and fall times of the PWM voltages, or often referred to as the fast dV/dt, subject the insulation system to more severe electrical stresses. A prominent example is the reflected wave phenomenon when the reflected voltage at the machine terminals can reach up to twice the DC bus voltage, or even up to four times in extreme cases. Such overvoltages have been reported to cause premature winding insulation breakdown. The insulation condition monitoring using effective PD detection becomes vital in such applications using VSDs. The fast dV/dt, nevertheless, causes significant switching noise in the traditional electrical PD detectors such as the coupling capacitors, crippling their capabilities to effectively detect PD, especially for medium- to high-voltage machines. Even if PDs can be successfully measured, the traditional PD pattern recognition techniques for electric machines, such as phase-resolved analysis, are no longer applicable because such techniques focus on the phase information of the PD occurrence with respect to the AC voltage. For pulsed voltages, the voltage switching plays a far more significant role than the phase of the reference voltage. This work strives to tackle the challenges in the detection and pattern recognition of PDs in electric machine insulation systems with pulsed voltage excitations. First a PD detection platform has been established to effectively detect and quantify PD events with pulsed voltage excitations. This platform employs nine sensors catering to five different PD physical manifestations, including electrical current, electromagnetic wave, optical light emission, acoustic ultrasound emission, and chemical ozone emission. PDs have been successfully and accurately measured for three types of samples: twisted pairs made of magnet wires, hand-wound stator coils, and professionally fabricated medium-voltage form-wound coils. Further a PD data processing procedure dedicated for data collected under pulsed voltages has been matured. Various machine learning algorithms and signal processing techniques are employed to implement feature extraction and pattern recognition from the collected PD data. The pattern unveiled facilitates the classification of PD pulses from different sources and locations. Moreover, the PD characteristics and patterns are investigated with pulsed voltage excitations of different parameters such as dV/dt. The effects of pulsed voltages, when compared with power-line-frequency AC voltage excitation, are hence demonstrated. Preliminary lifetime tests are performed on the twisted pair samples. The impact of recurring PDs on the lifetime of the low-voltage electric machine insulation system is further discussed in this dissertation.
Julia Zhang (Advisor)
241 p.
Electrical Engineering
partial discharge, partial discharge detection, electric machine insulation, pulsed voltage, variable speed drive, PWM voltage

Recommended Citations

Citations

  • Xiong, H. (2019). Detection and Pattern Recognition of Partial Discharge in Electric Machine Coils with Pulsed Voltage Excitation [Doctoral dissertation, Ohio State University]. OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=osu1555592971342072

    APA Style (7th edition)

  • Xiong, Han. Detection and Pattern Recognition of Partial Discharge in Electric Machine Coils with Pulsed Voltage Excitation. 2019. Ohio State University, Doctoral dissertation. OhioLINK Electronic Theses and Dissertations Center, http://rave.ohiolink.edu/etdc/view?acc_num=osu1555592971342072.

    MLA Style (8th edition)

  • Xiong, Han. "Detection and Pattern Recognition of Partial Discharge in Electric Machine Coils with Pulsed Voltage Excitation." Doctoral dissertation, Ohio State University, 2019. http://rave.ohiolink.edu/etdc/view?acc_num=osu1555592971342072

    Chicago Manual of Style (17th edition)

osu1555592971342072
472
© 2019, some rights reserved.Creative Commons License Detection and Pattern Recognition of Partial Discharge in Electric Machine Coils with Pulsed Voltage Excitation by Han Xiong is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported License. Based on a work at etd.ohiolink.edu.
This open access ETD is published by The Ohio State University and OhioLINK.