A Self-Sensing and Self-Powered Wearable System Based on Multi-Source Human Motion Energy Harvesting

Area of Science:

• Wearable technology
• Energy harvesting
• Biomedical engineering

Background:

• Wearable devices require continuous power, often limited by battery life.
• The human body generates significant wasted kinetic energy during movement.
• Existing energy harvesting methods for wearables face challenges in efficiency and integration.

Purpose of the Study:

• To develop a self-sensing and self-powered wearable system (SS-WS).
• To scavenge energy from waist motion and knee negative motion for powering devices.
• To integrate deep learning for motion recognition using harvested energy.

Main Methods:

• Utilized a three-degree-of-freedom triboelectric nanogenerator (TDF-TENG) for waist motion energy harvesting.
• Employed a negative energy harvester (NEH) to recover knee negative energy.
• Applied a Gate Recurrent Unit deep learning model to process triboelectric signals for motion recognition.

Main Results:

• The TDF-TENG, combined with deep learning, accurately recognized human motion states.
• The NEH achieved a power output of 27.01 mW at 8 km/h and a power density of 0.3 W/kg at 3 Hz.
• Demonstrated effective self-sensing and power supply for wearable devices.

Conclusions:

• The proposed SS-WS offers a viable solution for self-powered wearable systems.
• Integration of TENGs and NEH provides efficient energy harvesting from human motion.
• Deep learning enhances the self-sensing capabilities of the wearable system.