摘要
为了解决现有电力系统网络入侵检测方法漏检而导致检测效果不佳的问题,结合深度置信网络设计电力系统网络入侵检测过程。构建入侵检测模型,利用无监督学习方法对高维数据进行抽象化处理,保证特征向量全部映射到不同特征空间中。利用深度置信网络求解模型,达到实现全局最优的目的。使用深度置信网络训练入侵数据,采用反向传播算法计算受限玻尔兹曼机能量。快速学习测试集,构建入侵检测目标函数,获取每条测试数据的入侵类别。实验结果表明,该方法入侵检测波形与实际波形一致,且与实际数据存在最大为10类的误差,具有良好的检测效果。
the high⁃dimensional data to ensure that all feature vectors are mapped to different feature spaces.The depth confidence network is used to solve the model to achieve the purpose of global optimization.The depth confidence network is used to train the intrusion data,and the back propagation algorithm is used to calculate the limited Boltzmann machine energy.Quickly learn the test set,construct the intrusion detection objective function,and obtain the intrusion category of each test data.The experimental results show that the intrusion detection waveform of this method is consistent with the actual waveform,and there are up to 10 kinds of errors with the actual data,which has a good detection effect.
作者
左娟娟
陈宇民
朱红杰
彭文英
黄桂雪
ZUO Juanjuan;CHEN Yumin;ZHU Hongjie;PENG Wenying;HUANG Guixue(Baoshan Power Supply Bureau of Yunnan Power Grid Co.,Ltd.,Baoshan 678000,China)
出处
《电子设计工程》
2023年第24期85-89,共5页
Electronic Design Engineering
关键词
深度置信网络
电力系统
网络入侵
受限玻尔兹曼机
deep confidence network
power system
network intrusion
restricted Boltzmann machine
