Font Size: a A A

Research Of Network Intrusion Detection On Remote Monitoring System For Small Hydropower

Posted on:2015-01-04Degree:MasterType:Thesis
Country:ChinaCandidate:Y C WangFull Text:PDF
GTID:2298330467954981Subject:Computer software and theory
Abstract/Summary:PDF Full Text Request
Small hydropower is a clean, safe and renewable energy. It is the key point of the development direction of the national energy strategy. It also has important implications for environmental protection. At present, the information communicated between different power stations is increasingly delivered over the Internet. Remote monitoring system starts to be used in many hydropower stations. But various network security problems are also growing.Intrusion detection system is an important proactive network protection method. This paper analyses the intrusion detection technology research status and the concept of intrusion detection systems. This paper also applies machine learning technique to detect intrusion actions. The main contributions of this paper are as follows:(1) Propose a model analysis method to analyze six common machine learning algorithms. Propose a performance indicators included precision, recall, error rate, modeling time, sample complexity and the complexity of the algorithm. This performance indicator is used for comprehensive assessment of each machine learning algorithm. Experimental results show the model analysis process in detail and find out that the logistic regression is the optimal algorithm.(2) Propose a lightweight intrusion detection algorithm. This algorithm uses a well trained Gaussian anomaly detection model to filter the network traffic data and get the normal flow, malicious flow and undetermined flow. Then, the algorithm use logistic regression to make a final detection. According to the experiment results, this algorithm can filter nearly half of the network traffic data by using only two features. This can effectively save time and resources.(3) Design an intrusion detection system based on the proposed algorithm. The works include the system architecture design, database design and software development environment design. The proposed system provides an effective solution for the network intrusion detection of the small hydropower remote monitoring system.
Keywords/Search Tags:small hydropower, remote monitoring system, network intrusion detection, machinelearning
PDF Full Text Request
Related items