期刊
NEURAL PROCESSING LETTERS
卷 47, 期 2, 页码 601-618出版社
SPRINGER
DOI: 10.1007/s11063-017-9671-x
关键词
Memristor; Recurrent neural networks; Stability
资金
- Natural Science Foundation of China [61125303]
- Program for Science and Technology in Wuhan of China [2014010101010004]
- Program for Changjiang Scholars and Innovative Research Team in University of China [IRT1245]
- China Three Gorges University Science Foundation [KJ2013B020]
- Hubei Key Laboratory of Cascaded Hydropower Stations Operation and Control Program [2013KJX12]
- Hubei science and technology support program [2015BAA106]
- Yichang natural science research and application project [A15302a11]
Memristor is the fourth missing element. This paper discusses dynmacis memristive recurrent neural network with memristors as synapses. Firstly, it analyzes variation property of memristance under different external inputs with memristor simulation model. It concludes that memristance will be stable at one value if the direction of voltage is not changed and be varying periodically under periodically variable voltage. Next, it presents the memristive recurrent neural network model and gives local attractive region, one sufficient condition for memristive recurrent neural network under periodic voltage source. At last, an illustrative example is given for verifying our result.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据