基于大数据的配电网络复合攻击预测方法研究

发布时间：2018-01-01 18:28

本文关键词：基于大数据的配电网络复合攻击预测方法研究　出处：《电网与清洁能源》2016年10期 　论文类型：期刊论文

【摘要】：当前配电网络复合攻击环境较为复杂,有很大的不确定性,当前复合攻击预测方法收敛性较低,很难准确预测出攻击背景,导致预测结果不可靠。提出一种新的基于大数据的配电网络复合攻击预测方法,对配电网络中的报警数据进行预处理,以提高效率。在大数据分析的基础上,将隐马尔科夫模型应用于配电网络复合攻击预测中,要求配电网络依据得到的报警信息对隐藏的攻击行为进行挖掘。通过攻击行为概率分布、关联规则法得到初始状态矩阵、状态转移矩阵以及观察矩阵,引入粒子群算法对参数进行全局优化。通过隐马尔科夫模型中的Viterbi算法对配电网络复合攻击的攻击意图进行挖掘,预测下一步攻击。实验结果表明,所提方法具有很高的收敛性和预测精度。
[Abstract]:The current complex attack environment of distribution network is more complex, there is a lot of uncertainty, the convergence of the current complex attack prediction method is low, it is difficult to accurately predict the attack background. In order to improve the efficiency, a new compound attack prediction method based on big data is proposed to preprocess the alarm data in the distribution network in order to improve the efficiency. The hidden Markov model is applied to the compound attack prediction of the distribution network, which requires the distribution network to mine the hidden attack behavior according to the alarm information obtained. The probability distribution of the attack behavior is obtained. The initial state matrix, state transition matrix and observation matrix are obtained by association rule method. The particle swarm optimization (PSO) algorithm is introduced to optimize the parameters globally, and the Viterbi algorithm in Hidden Markov Model is used to mine the attack intention of the complex attack in the distribution network. The experimental results show that the proposed method has high convergence and prediction accuracy.
【作者单位】：国网江西省电力公司信息通信分公司;国网江西省电力公司;
【基金】：国家级青年基金面上项目(51677047)~~
【分类号】：TP311.13;TP309;TM73
【正文快照】： 行预处理,以提高效率。在大数据分析的基础上,将隐马尔科夫模型应用于配电网络复合攻击预测中,要求配电网络依据得到的报警信息对隐藏的攻击行为进行挖掘。通过攻击行为概率分布、关联规则法得到初始状态矩阵、状态转移矩阵以及观察矩阵,引入粒子群算法对参数进行全局优化。通

【相似文献】