论文部分内容阅读
为探究生物神经网络的自组织抗扰特性,进而为电子系统的电磁仿生防护提供新的思路和方法,本文基于神经网络中信息传递的机制和脉冲时间依赖突触可塑性(STDP)机制,探讨了突触可塑性与生物自适应特性的关系,然后选取Izhikevich神经元模型作为基本单元,以STDP机制调节的突触为桥梁,构建了四层的前馈神经网络模型,最后分析了该网络的自适应抗扰能力。仿真结果表明,基于STDP机制的神经网络具有很好的抗扰能力,且这种特性与STDP机制密切相关。基于此仿真工作将进一步进行领域转换,搭建模拟生物神经系统信息处理机制的神经元及突触的单元电路,并以该单元电路为基础设计具有自适应抗扰的电子电路。
In order to explore the self-organizing anti-disturbance characteristics of biological neural networks and to provide new ideas and methods for the electromagnetic bionics protection of electronic systems, based on the mechanism of information transmission and pulse-time-dependent synaptic plasticity (STDP) mechanism in neural networks, Synaptic plasticity and biological adaptive characteristics, and then select the Izhikevich neuron model as the basic unit to the STDP mechanism of the synapse as a bridge to build a four-layer feedforward neural network model, and finally analyzed the network adaptive Anti-interference ability. The simulation results show that the neural network based on STDP mechanism has good anti-interference ability, and this feature is closely related to the STDP mechanism. Based on this simulation work, further field transformation will be carried out to build neuron and synaptic cell circuits that simulate the information processing mechanism of biological nervous system, and based on the unit circuit, an electronic circuit with adaptive immunity is designed.