材料科学
触觉传感器
机器人
灵敏度(控制系统)
人工智能
接口(物质)
机器人学
计算机科学
瞬态(计算机编程)
感知
鉴定(生物学)
工作(物理)
色散(光学)
深度学习
联轴节(管道)
传输(电信)
触觉知觉
信号(编程语言)
机械臂
纳米技术
声学
纳米发生器
人工神经网络
传感器阵列
作者
Wenyan Qiao,Linglin Zhou,Xinyuan Li,Minghao Jiang,Zhihao Zhao,Jie Wang,Gui‐Bin Bian
摘要
Tribovoltaic nanogenerator (TVNG) offers self-powered, low impedance, and high direct current transient response, making it highly suitable for robotic sensing. However, the poor performance and lack of intelligent sensing terminals in the solid-solid TVNG of the contact separation mode have limited its advanced applications. We present a self-powered liquid TVNG sensor (LTVNGS) coupling sliding and contact-separation modes, integrated with deep learning for robotic tactile and gustatory dual-mode sensing system. The low-resistance aqueous solution at the metal-semiconductor interface promotes the charge transport efficiency of LTVNG, and the sensitivity has been improved from the nA to the µA level. The working mechanism of LTVNG was revealed, and the principle of enhancing sensitivity in the solid-solid noncontact LTVNGS through the qualitative and quantitative analysis of the GO dispersion solution was explained. Besides, LTVNGs achieve prediction of water droplet height. Furthermore, LTVNG achieves tactile material and gustatory liquid recognition with an average identification accuracy exceeding 98%, assisted by deep learning models through the perception of different materials and liquids. This work provides a scientific framework for multimodal TVNG and its application in intelligent robots, offers an innovative approach for robot liquid analysis, and enhances the ability to identify hazardous liquids and the HMI.
科研通智能强力驱动
Strongly Powered by AbleSci AI