Dual-Modified PDMS-Based Triboelectric Nanogenerator for Biomechanical Energy Harvesting and Deep Learning-Driven Caloric Consumption Calculation

Triboelectric nanogenerators (TENGs) offer significant potential for mechanical energy harvesting, but their development is constrained by the trade-off between high charge trapping capability and adequate electrical conductivity, which limits energy conversion efficiency. To tackle this issue, a dual-modification strategy involving the incorporation of polytetrafluoroethylene (PTFE) and multi-walled carbon nanotubes (MWCNTs) into polydimethylsiloxane (PDMS) was developed. PTFE was introduced to enhance charge trapping through its strong electron affinity, while MWCNTs formed a conductive network to facilitate rapid charge transport. Under optimized doping concentrations of 5 wt.% PTFE and 0.25 wt.% MWCNTs, the composite exhibited substantial performance enhancements, with open-circuit voltage, short-circuit current, and transferred charge increased by 294.43%, 521.0%, and 252.55%, respectively, compared to pure PDMS. A peak power density of 8.37 W/m² was achieved under a 20 MΩ load. The composite membrane exhibited a tensile strength of 4.21 ± 0.42 MPa, elastic modulus of 1.45 ± 0.07 MPa, and elongation at break of 169.80 ± 18.69%. The device displayed excellent thermal stability (20-100 °C) and slight UV-enhanced output, despite degradation at high humidity, it successfully retained 70.76% of its initial voltage at moderate humidity levels (55-65% RH). It maintained stable electrical output after 5000 seconds of continuous operation under mechanical stress, and generated clear voltage signals during various human motions such as walking, running, and jumping. Additionally, our motion recognition model based on deep learning techniques attained a high accuracy of 99.95%, demonstrating its effectiveness in action classification tasks. Combined with a UI visualization interface, it enabled the assessment of motion energy consumption. This study demonstrates an effective material design strategy for improving TENG performance and provides valuable insights for the advancement of high-efficiency wearable energy harvesting and information fusion systems.