Work

Development of a multi-coil wireless power transfer platform with deep learning-based free positioning

Year: 2025

Type: article

Abstract: This study presents a multi-coil wireless power transfer (WPT) platform that leverages deep learning to address the alignment limitations of conventional wireless chargers. To reduce hardware costs and improve coupling efficiency, a single-layer hexagonal array of transmitter coils is employed, enabling the primary side to achieve high coupling efficiency with fewer switches. The system design utilizes a Class-E amplifier combined with an LCC-S resonant compensation topology to minimize power loss and achieve stable resonance. To enhance positioning accuracy and reduce operational complexity, a deep neural network (DNN) model is introduced to predict the position of the secondary coil based on input current data from the primary side. Experimental results indicate that the proposed system achieves a positioning accuracy of 96.19%, with coil position detection taking less than 2 seconds, and an overall efficiency exceeding 70%, demonstrating the feasibility and superiority of the proposed approach.

Source: IEICE Electronics Express

DOI: 10.1587/elex.22.20250515

Authors: Yu-Lin Juan, Fu-Hsiang Ho, Yung-Lin Hsieh, Po-Hsuan Chen, Kai-Lun Hu

Institutions: National Changhua University of Education, National Formosa University

Language: English

Topic: Wireless Power Transfer Systems

Subfield: Electrical and Electronic Engineering

Field: Engineering

Domain: Physical Sciences

SDG: Industry, innovation and infrastructure

Open Access status: diamond

Funder: National Science and Technology Council