WiSARD无权重神经网络算法驱动的油田工控网络行为异常自动检测  

WiSARD Non-Weighted Neural Network Algorithm Driven Automatic Detection of Behavior Abnormal in Oilfield Industrial Control Networks

作　　者：丁楠 DING Nan(Digital Operation and Maintenance Center Measurement and Control R&D Testing Room of Daqing Oilfield First Oil Production Plant,Daqing,Heilongjiang 163001,China)

机构地区：[1]大庆油田第一采油厂数字化运维中心测控研发测试室,黑龙江大庆163001

出　　处：《自动化应用》2025年第6期129-131,共3页Automation Application

摘　　要：针对油田工控系统自动检测的不足,创新性地引入WiSARD无权重神经网络算法。该技术显著提升了网络行为异常的自动化检测能力。WiSARD算法通过高效提取网络行为特征,并应用主成分分析进行智能降维,有效精简了输入数据,进而训练模型,以精准识别异常行为。此技术不仅实现了检测过程的全面自动化,还显著增强了系统对复杂网络威胁的响应速度与准确性。实验证明,该自动检测方法能准确识别多种异常行为,有效提升检测准确率,满足油田工控网络日常检测需求。In response to the shortcomings of automatic detection in oilfield industrial control systems,an innovative WiSARD unweighted neural network algorithm is introduced.This technology significantly enhances the automated detection capability of behavior abnormal network.The WiSARD algorithm efficiently extracts network behavior features and applies principal component analysis for intelligent dimensionality reduction,effectively streamlining input data and training models to accurately identify abnormal behavior.This technology not only achieves comprehensive automation of the detection process,but also significantly enhances the system's response speed and accuracy to complex network threats.Experimental results have shown that this automatic detection method can accurately identify various abnormal behaviors,effectively improve detection accuracy,and meet the daily detection needs of oilfield industrial control networks.

关 键 词：神经网络 油田工控 特征提取 行为异常 自动检测 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]