Adaptive NFC WPT System Implementing Neural Network-Based Impedance Matching with Bypass Functionality

The work presents an adaptive impedance matching system controlled by an artificial neural network (ANN) for wireless power transfer (WPT) based on near-field communication (NFC) technology operating at 13.56 MHz. The system consists of two magnetically coupled resonators (MCRs) to establish wireless power transfer between a transmitter (TX) and receiver (RX). The MCR coils are based on class 1 standard NFC coils. In the TX, an adaptive impedance matching network (AIMN) is controlled by an ANN to optimize power transfer for changing TX-RX positions. Simulation and measurement data for changing TX-RX positions were used to train, validate, and test the ANN. A first analysis shows an improvement in the system’s power transfer efficiency for varying positions. However, the matching gain is not always overcoming the losses introduced by the AIMN, which then leads to bypassing the matching network. In contrast to previous work, this work analyses NFC-based WPT systems based on class 1 coils for the first time, giving valuable insight for future real-time adaptive NFC WPT systems.