基于多智能混合的网络性能管理系统研究-计算机应用技术专业论文.docxVIP

基于多智能混合的网络性能管理系统研究 基于多智能混合的网络性能管理系统研究 崔浩(计算机应用技术) 指导教师：全兆岐教授梁鸿副教授 捅芰 本文的主要工作是通过研究人工智能技术在网络性能管理中的应 用，发现目前的网络性能管理系统存在智能化不足的特点，为了提高 网络性能管理系统的智能化，通过研究专家系统、神经网络和粗糙集 理论三种技术，结合网络性能知识的特点，提出了一种基于多智能混 合的网络性能管理系统RSNNES，该系统采用了粗糙集理论与神经网 络、神经网络与专家系统松耦合的结合方式，粗糙集作为系统前端对 网络性能知识离散、约简和规则抽取，神经网络训练经属性约简后形 成的样本集，专家系统负责解释和完成人机交互。系统中粗糙集与神 经网络的融合是系统设计的重点，本文详细介绍了两者结合的原理， 并以此构建了系统的框架结构。本文还研究了网络性能的知识获取和 知识表示，实现了网络性能管理系统，提高了网络性能管理的智能化 水平。该系统对于处理其它领域问题也具有一定的适用性。 关键词：专家系统神经网络粗糙集网络性能管理 The The research of the network performance management based on the Multi—Intelligent Hybrid System CUIhao(ComputerApplicationTechnology) Directed by Professor TONG Zhao—qi，Associate-Professor L1ANG Hong Abstract n圮main idea of this paper is to study the intelligence of network performance management．The Multi—Intelligent Hybrid System was designed to improve the intelligence ofnetwork performance management by the integration ofvarious intelligent techniques． On the basis of improving the intelligence of network performance management，we designed the multi—intelligent hybrid system(RSNNES)’ by researching rough set．expert system and neural network．The way of integrating of the RSNNES is loose coupling by three techniques each other．111e knowledge of network performance is disposed，rule extracted， dispersed and reduced by rough set．The knowledge becomes the samples trained by USe of neural network．The expert system is responsible for interpreting mechanism and completing man—machine interface．The emphasis lies on the integration of ES and NN．The paper presents the principles of the integration of ES and NN and creates its structure．In addition，the paper illustrates the knowledge acquisition and representing about network performance，and completes the intelligence of network performance management．The variety of system Can applied to any other fields． Key Words：Expert System，Neural Network，Rough set，Network Performance Management 独创性声明零入声臻囊璺交豹论文是我个入在鼯爆撵霉下进行熬磷究工俦及 独创性声明 零入声臻囊璺交豹论文是我个入在鼯爆撵霉下进行熬磷究工俦及 取得的研究成果。尽我所知，除了文中