A High-Efficiency and Dual-Mode Huygens Rectenna Enabled Wireless Charging System for Supercapacitors Realized by Voltage-Dependent S-Parameter Analysis and Deep Learning

This paper introduces an efficient and dual-mode approach to wirelessly charge supercapacitors, as DC power suppliers for IoT devices, via Huygens rectenna. A novel voltage-dependent S-parameter analysis method is developed to maximize the far-field wireless power transfer (WPT) efficiency in supercapacitor-loaded Huygens rectenna systems. Thanks to the high-Q feature of Huygens rectenna, two charging modes, i.e., rapid and float, are realized by simply switching the operating frequency. Moreover, an innovative deep learning-assisted modeling method for supercapacitor is developed to accurately predict the charging time under different input RF powers. Operating within the 915MHz ISM band (902 to 928 MHz), the entire rectenna system is electrically small with the size ofka= 0.98. Experiment results show the highest average charging efficiency of 50.4% in the rapid charging mode at 904 MHz. Moreover, the frequency switching charging approach eliminates the need for additional charging management circuitry and effectively prevents the overcharging of supercapacitor under float mode. The developed method is further validated through a practical IoT sensor system. The system charges a 0.1 F supercapacitor in 433.2 seconds from 10 dBm input power, enabling over 3,400 seconds of operation. Experiment results demonstrate its viability for sustainable and continuous power supply.