Progress in artificial intelligence applications based on the combination of self-driven sensors and deep learning

In the era of the Internet of Things, developing a smart sensor system with sustainable power supply, easy deployment, and flexible use has become a challenging problem. Traditional power supplies, prone to frequent replacements or requiring charging during usage, hinder the advancement of wearable devices. The contact-to-separate friction nanogenerator (TEN G), composed of polychotomy ethylene (PTFE) and aluminum (AI) foils, addresses this issue by harvesting human motion energy through body movement arrangement. This energy is then utilized to monitor human motion posture by detecting changes in output electrical signals. In 2012, Academician Wang Zhonglin and his team invented the triboelectric nanogenerator (TENG), which operates as a self-driven sensor primarily powered by Maxwell displacement current. This enables direct conversion of mechanical stimulation into electrical signals during action, making it suitable for self-powered sensor applications. TENG-based sensors boast a simple structure and high instantaneous power density, providing a crucial tool for constructing intelligent sensor systems. Additionally, when combined with machine learning characteristics such as low cost, short development cycles, robust data processing, and predictive capabilities, the processing of the numerous electrical signals generated by TENG has a significant impact. The integration of TENG sensors is poised to drive the rapid evolution of intelligent sensor networks in various fields such as transportation, security, water conservancy, and construction, where urban sound management is essential. The method of seamlessly integrating sensors with Internet access is implemented using the NetBox network development platform, following these steps: 1) Initially, read the default parameter settings from pre-existing storage media, including parameters like IP address, subnet code, gateway address, sensor input and output types, and sensor range. These settings can be modified online to match the specific sensor used.2)Convert the sensor output signal into digital format using analog-to-digital conversion within the microcontroller.3)The microcontroller receives tasks from either the sensor or the Internet, such as medium, display, data processing, and web server, and processes them according to specified priorities. Existing sensor networks often comprise costly information sensing devices, limiting their large-scale deployment and functional measurement range due to cost and maintenance constraints.