Switched faults analysis of voltage source inverter (VSI) using short time fourier transform (STFT)

• Main Authors: Ahmad, Nur Sumayyah, Abdullah, Abdul Rahim, Bahari, Norhazilina, Abu Hassan, Mohd Arif
• Format: Article
• Language: English
• Published: Praise Worthy Prize 2014
• Online Access: http://eprints.utem.edu.my/id/eprint/13437/1/2014_IREMOS_Switches_Faults_Analysis_of_VSI_using_STFT.pdf; http://eprints.utem.edu.my/id/eprint/13437/; https://www.praiseworthyprize.org/jsm/index.php?journal=iremos&page=article&op=view&path%5B%5D=15451

Three-phase voltage source inverter (VSI) play a very important role in the present industrial life. The knowledge and information about faults characteristic of the VSI is important to prevent failure of equipment, data losses, high maintenance and economic losses. Therefore, to insure a continuous and safety operation of the VSI, switches faults analysis implemented to prevent further damages and the commonly open and short circuit faults are happened within the VSI. This paper present the analysis of switches faults VSI by using linear time-frequency distribution (TFD) which is short time Fourier transform (STFT) by simulation using MATLAB. STFT are appropriate technique to analyze non-stationary that consist of multi-frequency component and magnitude that represent signal in time-frequency representation (TFR). From the TFR obtained, parameters of the signals are estimated such as instantaneous average current, instantaneous of root means square (RMS) current, fundamental RMS current, total harmonics distortion (THD), total nonHarmonics distortion (TnHD) and total waveform distortion (TWD). According the analysis of switches faults parameters VSI is presented. Based on the signal parameters, characteristic of signal are calculated. The end of research, various characteristics of parameter VSI faults are clearly differentiate between open and short switches faults. The information of analysis switches faults is usefull to industrial application in the process for identifying fault detection and can fault can be identify early.