Multi-Objective Optimization of Circular Magnetic Couplers for Wireless Power Transfer Applications


Yilmaz T. , Hasan N., Zane R., Pantic Z.

IEEE TRANSACTIONS ON MAGNETICS, cilt.53, 2017 (SCI İndekslerine Giren Dergi) identifier identifier

  • Cilt numarası: 53 Konu: 8
  • Basım Tarihi: 2017
  • Doi Numarası: 10.1109/tmag.2017.2692218
  • Dergi Adı: IEEE TRANSACTIONS ON MAGNETICS

Özet

Recent expansion of emerging wireless power transfer technology has not been followed by an adequate systematic approach to the magnetic couplers design. The relation between design parameters (mutual inductance, coupling coefficient, leakage field flux density, coil quality factor, and so on) and structural parameters (coils and ferrite shape, size, thickness, and so on) has not been thoroughly explored, and practical designs often depend on trial-and-error methods supported by the finite-element modeling simulation to verify a final design. In order to fill that gap, this paper describes the adoption of multi-objective hybrid particle swarm optimization (MOHPSO) and multi-objective real-numbered particle swarm optimization (MORPSO) algorithms for a circular coupler design to increase performances and automate the design process. Objectives of this paper are threefold: 1) apply MOHPSO and investigate if there are some unconventional structures of circular couplers capable of outperforming traditional designs; 2) identify a minimum set of key optimization parameters for a circular magnetic coupler; and 3) demonstrate the operation of an MORPSO algorithm to optimize a circular magnetic coupler. The Pareto front concept is applied to provide multi-objective optimization. Two different multi-objective function pairs are considered, the first pair being the coil coupling coefficient (k) and maximum leakage field magnetic flux density (B-max); for the second pair, the product between quality factor (Q) and k is combined with (B-max). To validate the optimization algorithm effectiveness and accuracy, a selected magnetic coupler is fabricated and tested. Due to superior execution time, and satisfactory optimization capabilities, the MORPSO algorithm is employed to generate a final optimum design. Collected measurements demonstrate a very good agreement between the experimental and simulation results.