Experimental data validating the optimization of a wireless power transfer prototype employing a novel phase shift measurement system and frequency control

Resonant wireless power transfer (WPT) systems have been evolving and improving their designs over the last few years to efficiently charge electric vehicles, cellphones, and biomedical devices. In this article, we present to the scientific community the data obtained from the optimization of a resonant WPT prototype, operating at different vertical misalignments and load conditions, known to have an impact on the behavior of these type of systems. To maximize the power transferred to the load, we developed a proportional-integral frequency control algorithm that employs the phase-shift between the voltage and current waveforms in the transmitting antenna (resonance indicator) as a setpoint. Data on the performance and control optimization process of the prototype during laboratory tests were acquired using a LabVIEW interface, which was designed to capture information such as the evolution of the frequency, the phase-shift, and the load voltage, from multiple devices (a microcontroller, an oscilloscope, a digital multimeter, and a controllable power supply). The data were organized and presented in tables and graphs using Matlab. The importance of the datasets relies on the opportunity to utilize the information to model novel intelligent control algorithms, such as artificial neural networks controllers and adaptative neuro-fuzzy inference systems, which benefit from experimental training data.

谐振式无线电能传输 (Resonant Wireless Power Transfer, WPT) 系统近年来持续迭代优化设计，以实现电动汽车、移动电话及生物医学设备的高效充电。本文面向科研社群公开了一款WPT谐振样机的优化实验数据：该样机在不同垂直偏移量与负载工况下运行，而上述两类因素已被证实会对这类系统的运行特性产生显著影响。为最大化传输至负载的电能，本研究开发了一款比例积分型频率控制算法，该算法以发射天线处电压与电流波形间的相移（即谐振状态表征量）作为控制设定值。本研究通过LabVIEW（实验室虚拟仪器工程平台，Laboratory Virtual Instrument Engineering Workbench）开发的采集界面，采集了样机在实验室测试阶段的运行性能与控制优化过程数据：该界面可从微控制器、示波器、数字万用表及可控电源等多台设备中，获取频率变化、相移及负载电压等相关参数信息。实验数据通过Matlab整理为表格与图表形式进行呈现。本数据集的核心价值在于，其提供的实验数据可用于建模新型智能控制算法——例如人工神经网络控制器与自适应神经模糊推理系统，这类算法均可依托实验训练数据实现性能优化。