Optimum flux distribution with harmonic injection for a multiphase induction machine using genetic algorithms

Ayman Samy Abdel-Khalik, Shady Mostafa Gadoue, Mahmoud I. Masoud, Barry W. Wiliams

    Research output: Contribution to journalArticlepeer-review

    Abstract

    This paper investigates a nontriplen multiphase induction machine when fed with harmonic current injection with different sequences for an open loop optimized flux distribution that produce a quasi-square wave in the machine air gap. This maximizes iron utilization, giving more torque per ampere. The relation between the fundamental and other harmonic components can be determined for the best iron utilization using genetic algorithms where optimum flux distribution with different injected harmonic order can be obtained. This means, the target is to optimize the flux distribution during no-load to determine the optimum constants that guarantee approximate square wave air-gap flux. The paper focuses on an 11-phase machine that can be excited with harmonics up to the ninth. The technique is assessed using both winding function and finite element analysis methods. The prototype machine is fed from an 11-phase inverter. The system DSP control using genetic algorithm produces an optimum flux distribution by using winding sequence and harmonic current injection. Simulation results for the 11-phase dq model and prototype drive experimental results are presented.

    Original languageEnglish
    Article number5678812
    Pages (from-to)501-512
    Number of pages12
    JournalIEEE Transactions on Energy Conversion
    Volume26
    Issue number2
    DOIs
    Publication statusPublished - 1 Jun 2011

    Keywords

    • Eleven-phase
    • finite element analysis (FEA)
    • genetic algorithm (GA)
    • harmonic injection
    • multiphase induction machine
    • winding function

    Fingerprint

    Dive into the research topics of 'Optimum flux distribution with harmonic injection for a multiphase induction machine using genetic algorithms'. Together they form a unique fingerprint.

    Cite this