磺化木质素掺杂的动态配位水凝胶作为柔性传感器通过机器学习实现智能手势识别

文献信息

中文摘要

通过可穿戴电子设备将细微的人体运动转化为精确的数字指令，往往受到材料疲劳和信号噪声的阻碍，因此需要先进的水凝胶工程和智能数据解读。在本研究中，通过Fe3+配位交联和磺化木质素掺杂成功制备了一种多功能导电水凝胶。经优化的水凝胶具有高度互连的多孔结构和动态交联网络，表现出优异的力学性能和稳定的传感性能。该柔性传感器可贴附于人体，实现对关节运动、吞咽、眨眼和皱眉的实时监测，并具有出色的皮肤贴合性和信号稳定性。此外，基于标准摩尔斯电码设计了数字手势（0-9），并构建了一个两层长短期记忆（LSTM）网络（一种机器学习模型）用于手势识别。经过三倍数据增强后，该模型在测试集上达到了98.89%的识别准确率。这一集成策略为柔性可穿戴传感器在健康监测、人机交互和无声通信中的应用提供了一种新的技术解决方案。

英文摘要

The translation of subtle human motions into precise digital commands via wearable electronics is often hindered by material fatigue and signal noise, necessitating advanced hydrogel engineering and intelligent data interpretation. In this study, a multifunctional conductive hydrogel was successfully prepared through Fe3+ coordination crosslinking and sulfonated lignin doping. The as-optimized hydrogel features a highly interconnected porous structure and dynamic crosslinking network, demonstrating outstanding mechanical properties and stable sensing performance. The flexible sensor can be attached to the human body to realize real-time monitoring of joint movements, swallowing, blinking, and frowning, with excellent skin conformability and signal stability. Furthermore, digit gestures (0-9) were designed based on standard Morse code, and a two-layer long short-term memory (LSTM) network, which is a kind of machine learning model, was constructed for gesture recognition. After three-fold data augmentation, the model achieved 98.89% recognition accuracy on the test set. This integrated strategy provides a new technical solution for flexible wearable sensors in health monitoring, human-machine interaction, and silent communication.