基于神经网络的变电站告警信息安全态势智能评估方法
|
摘要
当前信息安全态势评估方法存在评估结果与实际值拟合度低、容错性能差、能耗高的问题,提出基于神经网络的变电站告警信息安全态势智能评估方法。利用消除模糊区域,扩大神经网络吸引范围,实现神经网络容错性分析。采用布谷鸟算法求取构建神经网络的最优权值和阈值,通过浮点数编码方式对布谷鸟进行编码,以神经网络的总误差值越小越好为依据,设计适应度函数,基于适应度函数的确定来设置布谷鸟种群规模和发现概率。通过编码和参数设定,初始化布谷鸟算法,根据迭代寻优完成最优权值和阈值求解。将历史告警信息安全态势值序列为当作输入数据输入至利用最佳权值和阈值构建的且已经训练好的神经网络,并将神经网络容错性分析结果引入信息安全态势评估中,实现变电站告警信息安全态势智能评估。实验结果表明,该方法评估准确性和容错性好、评估能耗低。
The current information security situation assessment method has the problems of low fitness,low fault tolerance and high energy consumption. A neural network-based intelligent assessment method for substation alarm information security situation is proposed. By eliminating the fuzzy region,the neural network attracting range is expanded,and the neural network fault tolerance analysis is realized. The cuckoo algorithm is used to obtain the optimal weight and threshold of the neural network. The cuckoo is coded by floating-point number coding. The fitness function is designed based on the fitness function based on the smaller the total error value of the neural network. The function is determined to set the cuckoo population size and probability of discovery. Through the coding and parameter setting,the cuckoo algorithm is initialized,and the optimal weight and threshold solution are completed according to the iterative optimization. The historical alarm information security situation value sequence is input as input data to the trained neural network constructed with the optimal weight and threshold,and the neural network fault tolerance analysis result is introduced into the information security situation assessment to realize the substation alarm. Intelligent assessment of information security posture. The experimental results show that the method has good accuracy and fault tolerance,and low energy consumption.
The current information security situation assessment method has the problems of low fitness,low fault tolerance and high energy consumption. A neural network-based intelligent assessment method for substation alarm information security situation is proposed. By eliminating the fuzzy region,the neural network attracting range is expanded,and the neural network fault tolerance analysis is realized. The cuckoo algorithm is used to obtain the optimal weight and threshold of the neural network. The cuckoo is coded by floating-point number coding. The fitness function is designed based on the fitness function based on the smaller the total error value of the neural network. The function is determined to set the cuckoo population size and probability of discovery. Through the coding and parameter setting,the cuckoo algorithm is initialized,and the optimal weight and threshold solution are completed according to the iterative optimization. The historical alarm information security situation value sequence is input as input data to the trained neural network constructed with the optimal weight and threshold,and the neural network fault tolerance analysis result is introduced into the information security situation assessment to realize the substation alarm. Intelligent assessment of information security posture. The experimental results show that the method has good accuracy and fault tolerance,and low energy consumption.
引文
[1]关兆雄,刘胜强,庞维欣.虚拟化技术在电力企业的移动生产应用研究[J].自动化与仪器仪表,2016,(2):5-6.
[2]田建伟,田峥,漆文辉,等.基于威胁传播的多节点网络安全态势量化评估方法[J].计算机研究与发展,2017,54(4):731-741.
[3]杨豪璞,邱辉,王坤.面向多步攻击的网络安全态势评估方法[J].通信学报,2017,38(1):187-198.
[4]王颖,王懿,陈文瑛,等.一种面向分布式分析的雾计算架构及其在电网安全风险评估中的应用[J].自动化与仪器仪表,2016,(9):128-130.
[5]王坤,邱辉,杨豪璞.基于攻击模式识别的网络安全态势评估方法[J].计算机应用,2016,36(1):194-198.
[6]顾兆军,王蕊莉.基于改进的模糊层次分析法的信息系统安全态势评估模型[J].计算机工程与科学,2016,38(10):2010-2017.
[7]文志诚,陈志刚,唐军.基于信息融合的网络安全态势量化评估方法[J].北京航空航天大学学报,2016,42(8):1593-1602.
[8]黄伟,刘琦,杨舒文,等.基于主动配电系统供电能力的安全态势感知方法[J].电力自动化设备,2017,37(8):74-80.
[9]龚俭,臧小东,苏琪,等.网络安全态势感知综述[J].软件学报,2017,28(4):1010-1026.
[10]丰颖,贠志皓,孙景文,等.输配协同的配电网态势快速感知方法[J].电力系统自动化,2016,40(12):37-44.
[11]赵冬梅,李红.基于并行约简的网络安全态势要素提取方法[J].计算机应用,2017,37(4):1008-1013.
[12]谢尧,吴柳,张思拓,等.基于大数据的电力通信网的安全防护系统及实现[J].电子设计工程,2017,25(19):131-135.
[13]吴果,陈雷,司志刚,等.网络安全态势评估指标体系优化模型研究[J].计算机工程与科学,2017,39(5):861-869.
[14]许晓燕.基于改进博弈模型的网络安全态势评估平台设计[J].现代电子技术,2016,39(12):87-90.
[15]王法玉,张晓洪.多源事件融合的网络安全态势评估方法[J].计算机工程与设计,2016,37(6):1440-1444.
[2]田建伟,田峥,漆文辉,等.基于威胁传播的多节点网络安全态势量化评估方法[J].计算机研究与发展,2017,54(4):731-741.
[3]杨豪璞,邱辉,王坤.面向多步攻击的网络安全态势评估方法[J].通信学报,2017,38(1):187-198.
[4]王颖,王懿,陈文瑛,等.一种面向分布式分析的雾计算架构及其在电网安全风险评估中的应用[J].自动化与仪器仪表,2016,(9):128-130.
[5]王坤,邱辉,杨豪璞.基于攻击模式识别的网络安全态势评估方法[J].计算机应用,2016,36(1):194-198.
[6]顾兆军,王蕊莉.基于改进的模糊层次分析法的信息系统安全态势评估模型[J].计算机工程与科学,2016,38(10):2010-2017.
[7]文志诚,陈志刚,唐军.基于信息融合的网络安全态势量化评估方法[J].北京航空航天大学学报,2016,42(8):1593-1602.
[8]黄伟,刘琦,杨舒文,等.基于主动配电系统供电能力的安全态势感知方法[J].电力自动化设备,2017,37(8):74-80.
[9]龚俭,臧小东,苏琪,等.网络安全态势感知综述[J].软件学报,2017,28(4):1010-1026.
[10]丰颖,贠志皓,孙景文,等.输配协同的配电网态势快速感知方法[J].电力系统自动化,2016,40(12):37-44.
[11]赵冬梅,李红.基于并行约简的网络安全态势要素提取方法[J].计算机应用,2017,37(4):1008-1013.
[12]谢尧,吴柳,张思拓,等.基于大数据的电力通信网的安全防护系统及实现[J].电子设计工程,2017,25(19):131-135.
[13]吴果,陈雷,司志刚,等.网络安全态势评估指标体系优化模型研究[J].计算机工程与科学,2017,39(5):861-869.
[14]许晓燕.基于改进博弈模型的网络安全态势评估平台设计[J].现代电子技术,2016,39(12):87-90.
[15]王法玉,张晓洪.多源事件融合的网络安全态势评估方法[J].计算机工程与设计,2016,37(6):1440-1444.