Optimal Control of LCL Tuned Inductive Power Transfer (IPT) Systems Using All Control Variables

Traditional control strategies for inductive power transfer (IPT) systems employing one, two, or three control variables cannot achieve optimal system efficiency with large variations in mutual inductance (
${M}$
) and output power (
${P}_{\mathrm {out}}$
) demands. Therefore, this article proposes a novel multivariable control strategy for LCL tuned bidirectional IPT (BD-IPT) systems. It uses all four possible control variables to allow the maximum possible efficiency to be tracked, even with large changes in
${M}$
and
${P}_{\mathrm {out}}$
. A mathematical model is employed to determine the load impedance conditions that give the maximum ac–ac efficiency. An extended time-domain model is employed to accurately establish zero-voltage switching (ZVS) operation. Its novel features include accounting for phase shifts due to frequency variations and for the coil equivalent series resistances (ESRs) of LCL tuned IPT systems. The proposed strategy is experimentally verified using a 1-kW prototype, and comparisons are made with traditional control strategies. In addition, the influence of the input to output dc voltage ratio on the characteristics of optimal control is investigated. The results show that the proposed strategy maintains higher efficiency than traditional strategies, given variations in
${M}$
and
${P}_{\mathrm {out}}$
.