Title :
Extending the operating range of the neuro-computing vector classification space vector modulation algorithm of three-level inverters into overmodulation region
Author :
Saeedifard, M. ; Bakhshai, A.R. ; Joos, G. ; Jain, P.
Author_Institution :
Dept. of Electr. & Comput. Eng., Isfahan Univ. of Technol., Iran
Abstract :
Three-level voltage source pulse width-modulated (PWM) inverters are used in high power industrial drive applications. Operating the converter in the undermodulation range of PWM technique restricts the drive operation in constant torque region. Therefore, an overmodulation strategy to extend the modulation region is essential. This paper is an extension to the operating range of the classification algorithm for the implementation of overmodulation schemes for SVM in three-level converters. Two schemes, depending upon the output voltage level, are developed to continuously increase the output voltage, in a controllable manner, up to the maximum attainable value. The technique requires less computational time, when compared with conventional SVM methods. The mathematical analysis is verified by means of simulation results.
Keywords :
PWM invertors; electric drives; electrical engineering computing; neural nets; PWM inverters; constant torque region; drive operation restriction; high power industrial drive; mathematical analysis; neuro-computing vector classification space vector modulation algorithm; operating range extension; output voltage level; overmodulation region; three-level inverters; three-level voltage source pulse width-modulated inverters; Classification algorithms; Pulse inverters; Pulse width modulation; Pulse width modulation converters; Pulse width modulation inverters; Space vector pulse width modulation; Support vector machine classification; Support vector machines; Torque converters; Voltage;
Conference_Titel :
Industry Applications Conference, 2003. 38th IAS Annual Meeting. Conference Record of the
Print_ISBN :
0-7803-7883-0
DOI :
10.1109/IAS.2003.1257575