摘要
为了解决传统的人工神经元电路难以连续调整和掉电存储权值的问题,使用由忆阻器构成的人工神经突触电路,设计出一种人工神经元电路.SPICE仿真实验结果证实了设计的可行性.所设计的神经元电路,在调整、存储权值方面相比于传统的人工神经元电路具有很大优势.
In order to solve the problem of continuous adjustment and power-down memory weights in traditional artificial neural circuits ,artificial neural circuits based on artificial synaptic circuits which is made of memristor is proposed in the paper .The proposed circuit is simulated by SPICE ,the results show that the circuit have a great advantage in terms of adjustment and memory weights compared with traditional artificial neural circuits .
出处
《微电子学与计算机》
CSCD
北大核心
2014年第8期171-176,共6页
Microelectronics & Computer
基金
湖北省自然科学基金项目(2011CDC075)
湖北省教育厅科研项目资助(D20101104)
