Highly hydrophobic MXene/PS@polypropylene fabric for human posture recognition assisted by machine learning
Abstract
In recent years, flexible wearable pressure sensors have emerged as a pivotal technology in the realms of intelligent health monitoring and artificial intelligence, steadily gaining traction as a prominent research focus. However, the conventional production process for conductive fillers is often cumbersome and costly, which limits their widespread application and large-scale manufacturing. In this study, a flexible pressure sensor based on polypropylene fluted woven fabric MXene/PS was proposed. The flexible pressure sensor uses spin-coated PS surface encapsulation technology to make the fabric surface hydrophobic, in order to improve the stability of the sensor. The sensor exhibits a wide strain range (0–565 kPa), excellent repeatability and stability (over 15 000 s), and a fast response–recovery time (75/159 ms), which can be attributed to the superior mechanical properties of the polypropylene-potted woven fabric. The MXene/PS pressure sensor can detect subtle deformations of small and medium-sized joints, making it suitable for detecting human motion signals. Additionally, combined with the deep belief network (DBN) algorithm, it can efficiently and accurately recognize human yoga postures, which shows its great application potential in human motion posture monitoring and low-cost flexible electronic products.