Artificial Touch Neurons Based on Starch Memristors and Hydrogel Sensors

摘要

The artificial touch neuron system has attracted the attention of researchers in the field of artificial intelligence because of its potential application in human health monitoring and human–computer interaction. A memristor can realize memory and computer integration, which is expected to overcome the limitations of the traditional von Neumann system. The combination of tactile sensors and memristors to construct artificial tactile neurons can realize highly parallel perception, memory, and processing of tactile information. In this paper, a biomemristor was fabricated using starch as the main material of the active layer. By embedding carbon nanotubes in the active layer of a starch memristor, the switching ratio of the device can be adjusted, and the device stability can be enhanced. A flexible haptic sensor based on starch was fabricated; the maximum strain could reach more than 15%, and the sensitivity could reach GF = 2.05, which could meet the needs of haptic perception. Artificial touch neurons can realize bionic synaptic functions such as paired-pulse facilitation, enhancement/inhibition, pulse time, and frequency-dependent plasticity. Tactile neurons realize the functions of memory, processing, and learning of tactile information and have significant application prospects in human–computer interactions, flexible robots, artificial intelligence, and other fields.