Empirical Regression-Based Model for Predicting Voltage and Current in a Household Wireless Power Transfer System

To power low-power household appliances, this study investigates the harvesting of RF signal energy using a Villard–Dickson rectifier circuit and a 470 MHz dipole antenna. The RF signal from the handy talkie is converted into DC voltage when measured from a distance of 5 to 25 cm. The voltage drops from 7.7V to 1.5V, and the current drops from 30mA to 10mA. A relationship showing the voltage and current decreasing almost inversely with distance was found thru logarithmic regression data analysis. A high coefficient of determination indicates compatibility with free-space propagation theory. At a distance of 5 cm, the highest RF-to-DC conversion efficiency reached 18.3%. The resulting regression model allows for performance prediction at distances beyond testing and serves as a practical guide for building wireless power transmission systems and low-power IoT applications. Additionally, this model also serves as a scientific reference for research and dissertations in the field of RF energy harvesting. Highlights: RF-to-DC conversion achieved up to 18.3% efficiency at 5 cm. Voltage and current decrease inversely with distance, supported by logarithmic regression. Regression model enables prediction and application in low-power IoT systems. Keywords: regresi logaritmik, rectifier Villard-Dickson, antenna dipole