神经形态工程学
计算机科学
晶体管
卷积神经网络
编码器
执行机构
压力传感器
人工智能
神经科学
材料科学
人工神经网络
电子工程
工程类
电气工程
电压
生物
机械工程
操作系统
作者
Sheng Li,Chengyao Yang,Liang Jia,Jun Zhang,Zihan Wang,Chen Kong,Weiyan Zhu,Qi Wang,Chang Guo
摘要
Drawing inspiration from the biomechanical and neural motor systems of trap-jaw ants, we introduce an artificial mechanoreceptive afferent nerve (AMAN) designed for tactile perception, with applications in neuromorphic computation and muscle actuation. The AMAN integrates a pressure sensor, spiking encoder module, IGZO synaptic transistor, flexible electromagnetic actuator, and spiking convolutional neural network to replicate key biological functions such as pressure perception, information processing, adaptive responses, and spatial pressure pattern recognition. This neuromorphic device exhibits a broad pressure perception range, rapid threshold response time (<4 s), and exceptional accuracy (99.3%) in spatially localized pressure mapping and handwritten digit identification. These advancements pave the way for bioinspired and biointegrated electronics that emulate biological neural systems.
科研通智能强力驱动
Strongly Powered by AbleSci AI