项目名称: 基于忆阻的神经网络动力学分析与应用研究
项目编号: No.61273200
项目类型: 面上项目
立项/批准年度: 2013
项目学科: 自动化技术、计算机技术
项目作者: 徐炳吉
作者单位: 中国地质大学(北京)
项目金额: 82万元
中文摘要: 本项目研究基于忆阻器的神经网络的动力学行为及其在动态联想记忆和混沌信号发生器中的应用。将结合忆阻器、神经网络、混合系统和混沌的综合优势,构建神经元模型,建立基于忆阻器的混合神经网络的稳定性、耗散性分析的准则和方法。解决忆阻器与混沌神经元及混沌神经网络的结合机制问题,分析基于忆阻器的混沌神经网络的动力学特性,解决动力学分析中出现的新问题,探讨针对忆阻器的性能的评价指标和分析方法。研究基于忆阻器的混沌神经网络在动态联想记忆和混沌信号发生器中的应用,设计电路板进行混沌信号发生器的实现。研究成果将进一步丰富神经网络理论,对新一代信息存储技术提供理论上的支撑。
中文关键词: 忆阻器;混合神经网络;混沌神经网络;动力学行为;
英文摘要: This project research on the dynamical behavior of memristor-based neural networks and the applications in dynamic associative memory and the chaotic signal generator.We will combine the advantage of memristor, neural networks, hybrid systems and chaos to construct the model of neurons, the criteria and methods for analyzing the stability, dissipativity of memristor-based hybrid neural networks.To solve the problem of the bonding mechanism of memristor and the chaotic neuron and the chaotic neural networks, analyze the dynamics of memristor-based chaotic neural network, solve new problems arising in the dynamics analysis, and to explore the evaluation indexes and analysis methods for the performance of memristor. Research on the application of memristor-based chaotic neural networks in a dynamic associative memory and chaotic signal generator, to design a circuit board for the realization of chaotic signal generator. The research will further enrich the neural networks theory, and provide theoretical support for a new generation of information storage technology.
英文关键词: Memristor;Hybrid neural network;Chaotic neural network;Dynamical behavior;