Voltage Source Inverter Switches Faults Classification And Identification Using Time-Frequency Distributions

• Main Author: Ahmad, Nur Sumayyah
• Format: Thesis
• Language: English; English
• Published: 2016
• Subjects: T Technology (General); TK Electrical engineering. Electronics Nuclear engineering
• Online Access: http://eprints.utem.edu.my/id/eprint/23311/; https://plh.utem.edu.my/cgi-bin/koha/opac-detail.pl?biblionumber=112185

Three-phase voltage source inverters (VSI) are utilized in a variety of industry applications.Although this technology has already achieved a certain level of maturity,due to their complexity and considering,three-phase VSI are often exposed to high stresses and unexpected faults may occur.Different types of faults occur in three-phase VSI such as open circuit,short circuit and gate misfiring that can influence reliability of entire system and disturb the performance.Hence, detection and classification of the three-phase VSI switches faults important for rectify failures and ensure the quality of power electronics system.This research presents the analysis of time-frequency distributions (TFDs) for three-phase VSI switches faults.The TFDs used are linear TFDs which are short time Fourier transform (STFT) and S-transform and bilinear TFD focusing on smooth-windowed Wigner-Ville distribution (SWWVD).The resulting time-frequency representations (TFRs) represent signals in the jointly time-frequency domains while the parameters of the signals are then estimated from the TFR.The signal parameters are instantaneous of root mean square (RMS) current,RMS fundamental current,average current,total waveform distortion (TWD),total harmonic distortion (THD) and total nonharmonic distortion (TnHD).From the signal parameters,the characteristics of the faults signals are calculated and are then used as input to a rule-based classifier to identify and classify the switches faults.The presented analysis is achieved by analyzing three-phase VSI for open and short circuit switches faults.In addition,based on the signal characteristics measurement,the best TFD is identified in terms accuracy,memory size and computation complexity used.Besides that,an experimental test also conducted to capture the real data for three-phase VSI switches fault using TFDs is reported to highligt the strength and weakness of each technique.Then,the best and effective technique for fault are discussed.The result shows that SWWVD is the best TFD technique for three-phase VSI switches faults detection and classification.