Flexible Gait Sensing and Machine Learning Recognition Based on Phase-Separated PVDF-HFP Films

Flexible wearable piezoelectric sensors have attracted increasing attention in human motion monitoring and motion classification applications due to their self-powered sensing capability and rapid response. In this work, poly(vinylidene fluoride-co-hexafluoropropylene) (PVDF-HFP) flexible piezoelectric films were fabricated using a phase separation method with different loading masses of PVDF-HFP to regulate the crystal structure and output signal characteristics of the films. X-ray diffraction and Fourier-transform infrared spectroscopy analyses demonstrated that an appropriate mass of PVDF-HFP promoted the formation of polar β-phase crystals, and the optimized film exhibited a β-phase content of 86.81%. The prepared films generated stable and distinguishable response signals under different gait conditions, indicating high potential for flexible motion sensing. Furthermore, machine learning-assisted motion classification was preliminarily performed based on the acquired sensing signals, achieving an accuracy above 90%. This work demonstrates the potential of phase-separated PVDF-HFP films for flexible gait sensing and wearable motion recognition applications.