通过机器学习实现基于摩擦纳米发电机的自驱动智能传感及其应用  

Self-powered sensing based on triboelectric nanogenerator through machine learning and its application

作　　者：张嘉伟 姚鸿博 张远征 蒋伟博 吴永辉[1] 张亚菊 敖天勇[3] 郑海务[1] Zhang Jia-Wei;Yao Hong-Bo;Zhang Yuan-Zheng;Jiang Wei-Bo;Wu Yong-Hui;Zhang Ya-Ju;Ao Tian-Yong;Zheng Hai-Wu(School of Physics and Electronics,Henan University,Kaifeng 475001,China;Key Laboratory of Artificial Micro-and Nano-structures of Ministry of Education,School of Physics and Technology,Wuhan University,Wuhan 430072,China;School of Artificial Intelligence,Henan University,Zhengzhou 475001,China)

机构地区：[1]河南大学物理与电子学院,开封475001 [2]武汉大学物理科学与技术学院,人工微结构教育部重点实验室,武汉430072 [3]河南大学人工智能学院,郑州475001

出　　处：《物理学报》2022年第7期378-402,共25页Acta Physica Sinica

基　　金：国家自然科学基金(批准号:52072111);河南省杰出青年科学基金(批准号:212300410004);河南大学一流学科培育项目(批准号:2019YLZDYJ04);河南省科技攻关项目(批准号:212102210025,212102210274)资助的课题。

摘　　要：在物联网时代,如何开发一种可持续供电、部署方便且使用灵活的智能传感器系统成为了亟待解决的难题..以麦克斯韦位移电流作为驱动力的摩擦纳米发电机(triboelectric nanogenerator,TENG)可直接将机械刺激转化为电信号,因此可作为自驱动传感器使用.基于TENG的传感器拥有结构简单、瞬时功率密度高等优点,为构建智能传感器系统提供了重要手段.同时,机器学习作为一种成本低、开发周期短、数据处理能力和预测能力强的技术,对TENG产生的大量电学信号处理效果显著.本文梳理了基于TENG的传感器系统通过采用机器学习技术进行信号处理和智能识别的最新研究进展,从交通安全、环境监测、信息安全、人机交互和健康运动检测等角度出发,概述了该研究方向的技术特点与研究现状.最后,深入讨论了该领域当前存在的挑战和未来的发展趋势,并分析了未来如何改进以期开拓更广阔的应用空间.我们相信机器学习技术与TENG传感器的结合将推动未来智能传感器网络的快速发展.In the era of The Internet of Things,how to develop a smart sensor system with sustainable power supply,easy deployment and flexible use has become an urgent problem to be solved.Triboelectric nano generator(TENG)driven by Max well’s Displacement Current can convert mechanical motion into electrical signals,thus it can be used as a self-powered sensor.Sensors based on TENGs have the advantages of simple structure and high instantaneous power density,which provide an important means to build intelligent sensor systems.Meanwhile,machine learning,as a technique with low cost,short development cycle,and strong data processing capabilities and predictive capabilities,is effective in processing the large amount of electrical signals generated by TENG.This article combines the latest research progress of TENG-based sensor systems for signal processing and intelligent recognition by employing machine learning techniques,and outlines the technical features and research status of this research direction from the perspectives of traffic safety,environmental monitor,information security,human-computer interaction and health motion detection.Finally,this article also in-depth discusses the current challenges and future development trends in this field,and analyzes how to improve in the future to open up a broader application space.It is suggested that the integration of machine learning technology and TENG-based sensors will promote the rapid development of intelligent sensor networks in the future.

关 键 词：物联网 摩擦纳米发电机 传感器 机器学习 

分 类 号：TM31[电气工程—电机] TP212[自动化与计算机技术—检测技术与自动化装置] TP181[自动化与计算机技术—控制科学与工程]