整流器(神经网络)
电容器
电感器
拓扑(电路)
能量收集
占空比
电气工程
能量(信号处理)
电压
物理
功率(物理)
计算机科学
工程类
人工智能
循环神经网络
随机神经网络
量子力学
人工神经网络
作者
Wenyu Peng,Xinling Yue,W.D. van Driel,Guoqi Zhang,Sijun Du
标识
DOI:10.1109/cicc60959.2024.10529024
摘要
A triboelectric nanogenerator (TENG) is a mechanoelectrical transducer promising to harvest ambient mechanical energy. A typical TENG model contains an alternating current source $(\mathrm{I}_{\mathrm{T}})$ in parallel with a small and time-varying capacitor $(\mathrm{C}_{\mathrm{T}})$ (Fig. 1 top-left), which introduces asymmetric electrical characteristics. Passive dual-output rectifiers (DORs) were proposed in [1] and [2], to tackle the asymmetric outputs, but they obtained low energy extraction performance due to energy loss on charging CT. Conventional active rectifiers were employed to improve performance, such as the bias-flip technique [3] [4]; however, a 1 's of mH large inductor is required to achieve good voltage flipping performance due to the high open-circuit voltage $(\mathrm{V}_{\text{OC}})$ of TENG. Although a capacitor-based bias-flip rectifier eliminates the use of an inductor [5], it was designed for piezoelectric transducers (PTs) and cannot be used in TENGs with varying $\mathrm{C}_{\mathrm{T}}$ . Besides, the maximum power point tracking (MPPT) for TENGs with active rectifiers has been seldom researched. A fractional open-circuit voltage (FOCV) technique was applied to triboelectric energy harvesting in [1] and [2], but measuring high Voc (hundreds of Volts) is very challenging. A duty-cycle-based (DCB) MPPT technique [6] was recently proposed which detects MPP without measuring any voltage; however, this technique was not suitable for TENGs with varying C T .
科研通智能强力驱动
Strongly Powered by AbleSci AI