Research On Key Technologies Of Data Security Risk Assessment For Active Distribution Networks

With the high integration of information and physical systems in the active distribution grid,the security threats to information systems will affect the physical systems.Risk assessment is an effective means to evaluate the security threats of distribution network information systems.The existing risk assessment mainly studies the threat degree of information security risks in distribution network information networks and systems,but it does not consider how the risk is transmitted between the power grid information system and the physical system,nor does it consider the impact of the risk in the information system on the physical system,and it cannot provide theory and method reference for the safety protection of the subsequent physical system of the power grid.Therefore,study the spread of risk in the information system in the power grid has very important research and practical significance for the transmission,generation and distribution of the power grid.This dissertation focuses on three aspects: the identification of vulnerable points of information systems for active distribution networks,the propagation of information security risks,and the assessment of information security risks.(1)Existing vulnerable point identification methods are mainly based on attack graphs,and all nodes are unified to identify vulnerable points,and this method cannot accurately and efficiently evaluate the nodes of all levels in the active distribution network.This dissertation proposes a method for identifying Vulnerability identification algorithm based on multi-layer attack graph network(VIA-MLAGN).In this dissertation,the data nodes in the active distribution network are divided into N layers according to the data control level,and multiple topology matrices are generated,each of which corresponds to an information layer.In addition,we consider the characteristics of too many nodes in the active distribution network.While considering the coverage of the attack path,the attack path is layered to optimize the repetition rate and complexity of the attack path.Experimental results show that optimizing the flow path of information node data in the active distribution network can improve the overall security of the active distribution network.(2)Most of the existing information risk assessment methods focus on the vulnerability assessment of the distribution network,without considering the interactive characteristics of the distribution network information collection devices.This dissertation combines the traditional infectious disease propagation theory,and proposes an information security risk propagation algorithm for distribution network based on the SEIR(susceptible-exposed-infected-recovered)infectious disease model(ISRP-SEIRIDM).The algorithm considers the information interaction between the information collection devices in the distribution network,and defines the connection nature and vulnerability of the security risks in the distribution network between the information collection devices.We also studied the influence of the number of information acquisition devices in the distribution network information system on the speed and maximum risk range of the risk transmission between distribution networks.The experiment analyzes the spread of risk in the distribution network information system.Experimental results show that optimizing the data flow of information nodes and optimizing the information transmission path in the active distribution network can reduce the range of risk propagation to some extent.(3)Existing risk assessment models do not take into account the impact of risk factors in information systems on physical systems.This dissertation proposes an information security risk assessment algorithm based on gene expression programming(ISRAA-GEP).Vulnerability identification algorithm and risk propagation algorithm are used to obtain the information security risk factors and their corresponding risk values in the information system of active distribution network.Combined with the idea of rough set,gene expression programming was used to mine the information security risk assessment function in active distribution network.The assessment model is used to predict the information security risk level in the active distribution network.Experimental results show that the risk assessment algorithm proposed in this dissertation can have a high accuracy in risk prediction.