Multiobjective Thermal Optimization Based on Improved Analytical Thermal Models of a 30-kW IPT System for EVs

Abstract

Thermal design is particularly important for high-power and compact inductive power transfer (IPT) systems having a limited surface area for heat dissipation. This article presents the thermal design and optimization of a 30-kW IPT system for electric vehicles. An improved analytical thermal model with high accuracy for liquid-cooled magnetic couplers was proposed by using the thermal network method (TNM). It considers heating components and thermal interface materials. Then, the multiobjective thermal optimization procedure of the liquid-cooled magnetic coupler was conducted with the presented model. Tradeoffs among temperature rise, weight, and cost were discussed, and an optimized solution was selected. The thermal FE models were established and compared with the thermal networks. Subsequently, the thermal performance of the system at different power levels and misaligned conditions was analyzed. The experimental setup based on the fiber Bragg grating sensors was built, and the prototypes were tested with an output power of around 28 kW. The error of stable temperature between the experiment and the prediction was less than 10% at the measurement points, verifying the thermal models. The proposed thermal models and optimization procedure accelerate the thermal design of IPT systems, toward higher power density.