Research On Security Monitoring And Information Attack Detection Strategy Of Power Terminal Based On Neural Network

With the deep integration of smart grid industrialization and informatization,communication protocols,equipment,systems have gradually moved from isolation and closed to inclusive and open.The large number of access terminals in the power grid system,the complex access environment,and the widespread use of wireless communication methods make the communication process face security threats,which directly affect the normal operation of the smart grid.The on-site power distribution terminal is mainly connected to the power distribution automation system through optical fiber and wireless network.Due to the relatively weak security protection measures and the enhancement of hacker attack methods,and the distribution terminal has a somewhat multi-faceted distribution feature,attackers can maliciously damage the power distribution master station system by forging terminal identities and replay attacks,which in turn affects the safe and reliable power supply of the power distribution system to users.In the research process of this paper proposes a power terminal safety monitoring technology based on deep learning,which aims to protect the power terminal at the equipment level and network level.First,a device-level security protection strategy is proposed.By collecting bypass information of power terminal,and then using LSTM neural network to capture the timing changes of terminal power consumption,the device-level security monitoring of power terminals is realized.The safety monitoring method based on bypass signal proposed in this paper realizes the non-intrusive monitoring of power terminal,which can effectively monitor the power terminal without affecting the normal operation of the power terminal.Secondly,it proposes a network-level security protection strategy,collects the flow message information and implements information attack detection on the power terminal at the network level through methods such as data cleaning,feature extraction,and model training.Then the power terminal conducts safety monitoring.This paper verifies the feasibility of the proposed method by building a physical simulation platform.The experimental results show that the strategy proposed in this paper can improve the safety of the power terminal.