Hao Zhang,Dongzhi Zhang,Ruiyuan Mao,Lina Zhou,Chunqing Yang,Yan Wu,Yukun Liu,Yuncheng Ji
Self-powered sensing technology and smart perception technology have broad application prospects in flexible and wearable electronics . In this work, a flexible triboelectric nanogenerator (RMP-TENG) based on a room-temperature vulcanized silicone rubber (RTV)@(Molybdenum disulfide (MoS 2 )/Polyvinyl chloride (PVC)) functional layer is developed using a layer-by-layer self-assembly and material doping strategy. The RTV@(MoS 2 /PVC) functional layer is divided into a charge-generating layer (RTV/PVC) and a charge-trapping layer (RTV/MoS 2 ). The synergistic effect of PVC and MoS 2 has improved the surface roughness and charge transfer efficiency of RMP-TENG, doubling its output performance to 1055 V and 112 μA. In order to further improve the tactile sensing accuracy of RMP-TENG, a deep learning model based on Convolutional Neural Network (CNN) and Gated Recurrent Unit (GRU) is developed. It predicts the type of contact material based on the features of tactile signals, achieving a prediction accuracy of 93.975%. Additionally, by combining mobile smart terminals, the CNN-GRU model, and RMP-TENG, a wireless access control system based on self-powered tactile sensing and intelligent material recognition is developed. Through the optimization of these experiments and algorithms, RMP-TENG has achieved real-time material recognition capabilities. This demonstrates the broad application prospects of RMP-TENG in wearable energy supply, intelligent sensing, human-computer interaction, and other areas.
Currently, deep-learning-assisted triboelectric nanogenerators (TENGs) have shown great potential for human-computer interaction. The Ecoflex-based polyvinyl-alcohol layer (Ecoflex/PVA) and the Ecoflex-based graphitic carbon nitride layer (Ecoflex/g-C 3 N 4 ) are sequential spin-coated on the substrate of flexible cotton yarn . The Ecoflex/g-C 3 N 4 and Ecoflex/PVA based silicone composite layer (SCL@(g-C 3 N 4 /PVA)) is designed as a high-output artificial skin in TENG. The flexible silicone composite layer plays a crucial role in enhancing output of the TENG. The silicone composite layer-based triboelectric nanogenerators (SCL-TENG) with 10 wt% PVA and 1.6 wt% g-C 3 N 4 yielded the optimal output (720 V, 134 μA, 0.255 mW/cm 2 ) under the pressure of 5 kPa and frequency of 8 Hz. By applying the Internet of Things technology, the single electrode mode SCL-TENG can be integrated into the intelligent sensing system to control and monitor electronic and electrical systems . In addition, the single electrode mode SCL-TENG is capable of sensing and distinguishing the instantaneous mechanical contact generated by balls with different materials, which can be high-accuracy identified with the assistant of deep-learning method of convolutional neural network-gate recurrent unit (CNN-GRU). It shows that the flexible silicone composite layer has great potential in the next generation of AI and intelligent interactive applications.
Artificial neural networks based on artificial synaptic thin-film transistors (ASTFTs) offer promising opportunities for efficient and scalable information processing. However, implementing both spatial and temporal processing within a unified system requires complex hybrid network architectures, placing high demands on the tunability of individual synaptic devices. Here, we report a reconfigurable photosensitive synaptic transistor composed of FAPbI₃ colloidal quantum dots (CQDs) and an InO x channel, integrated with a modulation scheme tailored for hybrid convolutional-sequence neural computing. The device enables mode-specific plasticity through programmable optical and electrical stimuli. For spatial signal processing, it achieves a long-term potentiation (LTP) linearity of 0.062, long-term depression (LTD) linearity of 0.143, and an asymmetry ratio (AR) of 0.185. For temporal learning, the dynamic fading memory (FM) time constant is tunable from 42 to 75 ms, enabling dynamic short-term memory (STM) behavior. Utilizing these synaptic features, we construct a hybrid CNN-GRU neuromorphic system that achieves 94.2% accuracy in real-time gesture recognition within only 7 training epochs, using a custom dataset over 200 times larger than conventional benchmarks.This work demonstrates a unified neuromorphic hardware platform capable of spatiotemporal cognition, offering new strategies for the development of intelligent optoelectronic systems.
危险品化学品经营许可证（带存储）