小水电远程监控系统网络入侵检测研究

发布时间：2018-06-25 06:43

本文选题：小水电 + 远程监控系统　；参考：《浙江工业大学》2014年硕士论文

【摘要】：小水电是一种清洁、安全的可再生能源,对于环境保护有重要意义,是国家能源发展的重点战略方向。目前,电站间的通信逐渐开始互联化,远程监控系统开始运用于各种类型的水电站,随之而来的网络安全问题也日趋增多。入侵检测系统是一种主动发现网络异常的安全防护技术,对于保障计算机网络安全十分重要。本文在分析入侵检测技术研究现状的基础上,研究基于机器学习技术的入侵检测方法及系统构建,选择了适合小水电环境下的检测方法,具有重要的研究价值。主要工作如下： (1)提出小水电环境下机器学习算法对入侵行为识别效果的评估指标,能够综合评估算法的准确率、召回率、误差率、建模时间、样本复杂度以及算法本身的复杂度。通过六种常用的机器学习算法进行了验证,实验详细表述了算法评估的每一个过程,最后得到最佳算法,逻辑回归。 (2)提出了一种基于高斯异常检测和逻辑回归模型的轻量级入侵检测算法,首先使用训练的高斯异常检测模型对流量数据进行过滤,分别得到正常流量、异常流量和待定流量,然后使用逻辑回归算法对待定流量再进行最后判断。实验结果显示,该算法仅使用描述网络流量行为的2个特征,即可对将近一半的网络流量进行有效的判断,节省了检测时间以及硬件计算资源。 (3)基于上述成果,开发了小水电远程监控系统网络入侵检测系统,包括了系统架构设计、数据库设计和软件开发环境设计,为小水电远程监控系统网络入侵检测提供了有效的解决方案。
[Abstract]:Small hydropower is a kind of clean and safe renewable energy, which has important significance for environmental protection and is the key strategic direction of national energy development. At present, the communication between power stations is becoming interconnected gradually, remote monitoring system has been applied to all kinds of hydropower stations, and the network security problems are also increasing day by day. Intrusion detection system (IDS) is a kind of security protection technology which can actively detect network anomalies. It is very important to ensure the security of computer network. On the basis of analyzing the present situation of intrusion detection technology, this paper studies the intrusion detection method and system construction based on machine learning technology, and selects the detection method suitable for small hydropower environment, which has important research value. The main work is as follows: (1) the evaluation index of machine learning algorithm for intrusion recognition in small hydropower environment is proposed, which can comprehensively evaluate the accuracy, recall rate, error rate and modeling time of the algorithm. The complexity of the sample and the complexity of the algorithm itself. Six commonly used machine learning algorithms are used to verify the algorithm. The experiment describes each process of algorithm evaluation in detail. Finally, the best algorithm is obtained. (2) A lightweight intrusion detection algorithm based on Gao Si anomaly detection and logical regression model is proposed. Firstly, the trained Gao Si anomaly detection model is used to filter the traffic data to obtain the normal traffic. The abnormal flow and undetermined flow are determined by the logic regression algorithm. The experimental results show that the proposed algorithm can effectively judge nearly half of the network traffic using only two features describing network traffic behavior, and saves detection time and hardware computing resources. (3) based on the above results, The network intrusion detection system of small hydropower remote monitoring system is developed, including system architecture design, database design and software development environment design, which provides an effective solution for small hydropower remote monitoring system network intrusion detection.
【学位授予单位】：浙江工业大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP393.08

【参考文献】