Abstract
In this study, a flexible hydrogel film sensor based on the intermixing of poly(vinyl alcohol) (PVA) and biomass-derived carbon nanoparticles (CNPs) was prepared and microstructures were constructed by replicating sandpaper templates on its surface. The sensor thus has good overall sensing performance with a sensitivity of 101 kPa(-1), a fast response/recovery time of 22 ms and 20,000 fatigue cycles. The sensor was experimentally verified to accurately capture human joint movements, current signals of written letters, and weight differences in the size of spherical objects. Based on this, a breathing phase classification framework was constructed using the 1D-CNN algorithm, achieving a synergistic enhancement effect between environmentally scalable materials and Deep learning algorithms. This approach not only improves the signal discrimination function, but also provides new ideas for wearable medical monitoring, haptic feedback and intelligent robot human-machine interface.