基于石墨烯-PEDOT:PSS/聚对二甲苯复合薄膜的柔性皮肤电极在手势识别中的应用

Application of Flexible Skin Electrodes Based on Graphene-PEDOT:PSS/Parylene Composite Films in Gesture Recognition

  • 摘要: 表面肌电(Surface Electromyography, sEMG)电极是连接人体肌肉活动与外部分析系统的关键“桥梁”。为克服传统Ag/AgCl凝胶电极在长期可穿戴表面肌电监测中存在的凝胶易干、皮肤刺激及机械顺应性不足等局限,本研究制备了一种基于石墨烯-PEDOT:PSS/聚对二甲苯(Graphene-PEDOT:PSS/Parylene, GPP)复合体系的高性能柔性干电极。通过引入十二烷基硫酸钠与双三氟甲烷磺酰亚胺锂优化PEDOT:PSS的导电性与石墨烯界面相容性,并利用石墨烯与PEDOT:PSS间的π-π堆叠作用构建三维导电网络,成功制备出兼具优异柔性、高导电性(面电阻低至2.62 Ω/sq)及良好机械稳定性的GPP薄膜电极。该电极在动态弯曲及重复粘附-剥离测试中电学性能稳定,且其电极-皮肤界面阻抗在100 Hz下仅为8.8±0.7 kΩ,显著低于商用Ag/AgCl电极(25.1±1.6 kΩ)。在sEMG信号采集中,GPP电极表现出高达26.47 dB的信噪比,远优于Ag/AgCl电极(11.98 dB)。进一步地,将基于手势动作的sEMG信号经预处理后输入一维卷积神经网络模型,实现了对六种手势高达95.83%的平均识别准确率。本研究为开发适用于长期、高精度生物电信号监测的柔性可穿戴电子器件提供了有效的材料策略与技术方案,在康复医学、人机交互及智能运动监测等领域具有广阔应用前景。

     

    Abstract: Surface electromyography (sEMG) electrodes serve as the crucial "bridge" connecting human muscle activity with external analysis systems. To overcome the limitations of traditional Ag/AgCl gel electrodes, such as gel drying easily, skin irritation, and insufficient mechanical compliance, in long-term wearable surface electromyography monitoring, this study has developed a high-performance flexible dry electrode based on the graphene-PEDOT:PSS/polyarylene (Graphene-PEDOT:PSS/Parylene, GPP) composite system. By introducing sodium dodecyl sulfate and lithium difluoromethanesulfonimide to optimize the conductivity of PEDOT:PSS and the interface compatibility with graphene, and utilizing the π-π stacking interaction between graphene and PEDOT:PSS to construct a three-dimensional conductive network, a GPP film electrode with excellent flexibility, high conductivity (with a surface resistance as low as 2.62 Ω/sq), and good mechanical stability was successfully prepared. This electrode exhibited stable electrical performance during dynamic bending and repeated adhesion-deposition tests, and the electrode-skin interface impedance was only 8.8±0.7 kΩ at 100 Hz, significantly lower than that of commercial Ag/AgCl electrodes (25.1±1.6 kΩ). In sEMG signal acquisition, the GPP electrode showed a signal-to-noise ratio of up to 26.47 dB, far superior to that of Ag/AgCl electrodes (11.98 dB). Further, by preprocessing the sEMG signals based on gesture actions and inputting them into a one-dimensional convolutional neural network model, an average recognition accuracy of up to 95.83% for six gestures was achieved. This study provides an effective material strategy and technical solution for developing flexible wearable electronic devices suitable for long-term, high-precision biological signal monitoring, and has broad application prospects in rehabilitation medicine, human-computer interaction, and intelligent motion monitoring.

     

/

返回文章
返回