الفهرس 
Chapter 1Only 14 pages are availabe for public view
 |  | from | 195 |  |  |
| | | from | 195 | | |
Abstract
”Power electronic converters fed three-phase induction motor drive systems are very susceptible to suffer from critical failures such as open switch fault of the power electronic converter and internal stator faults of the machine. Monitoring and early fault detection of such faults improve the reliability and ensure safe continuous operation of the motor drive system. This thesis presents four distinct methods for diagnosing faults within induction motor drive systems. These methods utilize various currents signal processing techniques to identify and classify common faults such as open switches, open phase, malfunctioning sensors and inter-turn short circuit faults. The first technique employs the analysis of current signals over a single cycle. By identifying the maximum and minimum points within the signal, this method enables the diagnosis of open switch and open phase faults. The second method utilizes the αβ rotating reference frame currents for fault diagnosis. Here, normalized extracted DC components of currents are employed to identify open switch faults within the motor drive system. The third method leverages the ratio between extracted DC components from the line currents and their absolute along with the DC component of absolute line currents and absolute line currents sum. This approach allows the detection of 33 different faults types including open switches, open phases, and broken current sensor faults. A fuzzy logic-based subsystem is dedicated to accurate reporting of different fault scenarios. The fourth method utilizes normalized DC components of three line currents alongside normalized, filtered, negatively oriented motor line current components in the dq reference frame. This combined analysis offers the capability to diagnose both open switch faults and inter-turn short circuit faults, even short faults with small turns number percentage. Additionally. This method develops a technique to accelerate detection in case of short faults with high turns number percentage. All of the proposed four methods are verified through simulation using MATLAB/Simulink as well as experimental setup underdifferent fault scenarios and loading conditions. The obtained results proved the reliability and efficacy of the proposed methods under different fault cases.