Research on Abnormal Detection Technology of Real-Time Interaction Process in New Energy Network
Bo Peng, Yanyan Wang, Xin Yan Li, Junhui Cai, Jiaxuan Fei, Chen Wei · 2019
Efficient and accurate detection of abnormal interaction information at network-related terminals of new energy plants and stations can effectively improve the network security and protection capabilities of new energy plants and stations. Firstly, the network attack scenarios, malformed messages and irregular business instructions of new energy plants and stations are analyzed. Secondly, the deep parallel parsing technology of real-time interactive protocol for new energy plants and stations is proposed to improve the efficiency of protocol parsing. Thirdly, a real-time interactive process anomaly detection technology based on feature matching is proposed. K-NN classification algorithm is used to match the characteristic vectors of network data packets in the new energy plant and station system, which realizes the anomaly detection of network attack scenarios, malformed messages and irregular business instructions in the new energy plant and station system. Finally, a simulation experiment environment of the new energy plant and station system is built to verify the method proposed in this paper. The experimental results show that the algorithm has high ability of anomaly detection and low false alarm rate. It is of great significance to improve the level of network security protection of new energy plants and stations, and to ensure the safe and stable operation of new energy plants and stations.