An Artificial-Intelligence-Assisted Optimization of Imperceptible Multimode Rectenna

In this letter, we propose and experimentally demonstrate compact, low-profile, and optically transparent antennas for multiband and multirange wireless power transfer (WPT) applications. Specifically, we put forward new types of transparent multiband antennas that can perform the near-field reactive WPT (13.56 MHz), as well as the far-field radiative WPT (980 MHz and 2.45 GHz) within a single device. Furthermore, such an antenna is integrated with compact frequency-scalable rectifying circuits to form an unseeable multimode WPT device. We show that a hybrid inductive (13.56 MHz) and radiative (980 MHz and 2.4 GHz) WPT device can be realized with a modified inverted-F antenna structure connected to spiral coil virtual ground. To meet the stringent design requirements of this unobtrusive multiband antenna, a state-of-the-art machine-learning-assisted global optimization method (parallel-surrogate-model-assisted hybrid differential evolution for antenna optimization) is exploited for global optimization. We envision that the proposed transparent and flexible WPT and energy harvesting devices can be beneficial for many applications, including ubiquitous wireless charging based on smart windows and glasses, solar radio frequency integrated power supply, wearable or textile electronics, and Internet of Things.