首页 | 本学科首页   官方微博 | 高级检索  
     检索      

模仿神经元网络抗扰特性的电磁防护仿生研究
引用本文:满梦华,蔡娜,马贵蕾,王震.模仿神经元网络抗扰特性的电磁防护仿生研究[J].装备环境工程,2017,14(4):9-15.
作者姓名:满梦华  蔡娜  马贵蕾  王震
作者单位:1. 军械工程学院 静电与电磁防护研究所,石家庄,050003;2. 军械技术研究所,石家庄,050003;3. 廊坊市健康教育研究所,河北 廊坊,065000
基金项目:国家自然科学基金项目(51407194)
摘    要:目的研究神经元在噪声干扰环境下信息处理的抗扰特性,为电磁防护仿生研究提供有益借鉴。方法利用Hodgkin-Huxley模型建模神经元电信号的产生,结合S空间编码理论分析神经信息的表达。在此基础之上,研究神经信息处理在噪声干扰环境下的抗扰特性。建立具有噪声耦合方式的神经元数学模型,并在不同噪声强度下,计算神经元输出电信号对输入刺激的S空间编码,讨论噪声对编码的影响。结果在S空间中,神经元将输入刺激信号编码成符号序列,符号序列间的排序关系与输入信号频率间的排序关系所对应。输入噪声能够改变符号序列的值,但并没有改变符号序列间的排序关系,从而不会影响神经元在S空间中所表达的信息。结论 S空间编码是神经元抵御输入噪声干扰的一种重要机制,值得电子系统借鉴,以提高其抗扰能力。

关 键 词:S空间编码  Hodgkin-Huxley模型  噪声  抗扰
收稿时间:2016/10/31 0:00:00
修稿时间:2017/4/15 0:00:00

Study on Electromagnetic Protection Bionics by Mimicking the Anti-interference Me-chanism of Neural network
MAN Meng-hu,CAI N,MA Gui-lei and WANG Zhen.Study on Electromagnetic Protection Bionics by Mimicking the Anti-interference Me-chanism of Neural network[J].Equipment Environmental Engineering,2017,14(4):9-15.
Authors:MAN Meng-hu  CAI N  MA Gui-lei and WANG Zhen
Institution:Electrostatic & Electromagnetic Protection Institute, Ordnance Engineering College, Shijiazhuang 050003, China,Ordnance Technical Research Institute, Shijiazhuang 050000, China,Electrostatic & Electromagnetic Protection Institute, Ordnance Engineering College, Shijiazhuang 050003, China and Langfang Institute of Health Education, Langfang 065000, China
Abstract:Objective To study the good anti-interference ability of neural system of organism appears during infor-mation process, which can bring enlightenment to the study of bio-inspired electromagnetic protection.Methods We study the underlying mechanism of neural information processing in noise by using the modified bursting Hodg-kin-Huxley neuron model to construct simulation models of neural system and S-space coding theory to analyzing neural information. The neural simulation model with different noise intensity is built, the neural information is coded by S-space coding theory, and influence of noise on neural coding is discussed.Results The results show that the neural in-formation is encoded to symbol sequences in S-space and the frequency of input signal has monotonic relationship with the symbol sequences. The input noise changes the symbols of the symbol sequences but does not change the monotonic relationship, that is, the input noise doesn't influence the information processing in S-space.Conclusion S-space coding theory is an important mechanism for anti-interference ability of neural system, which is worth to draw lessons from by electronic system.
Keywords:S-space coding theory  Hodgkin-Huxley model  noise  anti-inference
本文献已被 CNKI 万方数据 等数据库收录!
点击此处可从《装备环境工程》浏览原始摘要信息
点击此处可从《装备环境工程》下载免费的PDF全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号