| With electric power information technology and communication network gradually expanded, the number of equipments and the workload of power system increases rapidly. Information room is a data storage, transmission, management place, the security and stability of the information room plays an important role at the operation of electirc power communication network. The information room monitoring and security evaluation can help operation staff know about the current information room’s equipment running status, internaml power and environment dynamic status information, through the the identification of abnormal data, master the information room security risks exist, and evaluate the security of whole room, to make a rational decision has important significance.This paper conducted the research on the information room security operations from two aspects, one is the anomaly identification method of monitoring data of information room, the other is a fuzzy security evaluation model of information room.In the aspect of anomaly identification method of monitoring data of information room, information room monitoring exists following problems: information room monitor data mode and type are huge and different, it is different to direct construction of the measurement model. The anomaly data quantity is less, different types of information of each monitoring data is difficult to identify the anomaly data by setting a simple threshold. It is also need rapid identification of anomaly data in the monitor data. In the view of the above problmes, this paper presented boundary outlier of normal cluster anomaly identification method. Firstly, the normal clusters and the corresponding boundary samples are obtained by the improved CURE clustering algorithm. Then, the anomaly data identification algorithm based on boundary samples is designed. It has ability of handling excptions in real and history data. Through the experiment analysis, the anomaly identification method is fast and accurate, and it provides a feasible mothod for the anomaly identification and hazard investigation in the information room security operations.In the aspect of information room fuzzy security evaluation, according to the information room disaster events is uncertain and the influence factors are fuzzy, and the vulnerability caused by difficulty analysis of collaborative information heterogeneity, and there are many factors influencing the evaluation results, proposed a fuzzy security evaluation model based on optimal clustering conditions. Using the optimal clustering conditions on evaluation data set to obtain fuzzy division set, at the same time confirme the fuzzy membership function and weights of each index, at the end build the two level security evaluation model based on multi factor, to strengthen the effectiveness of security evaluation. Through the experimental analysis, the model can effectively evaluate the security status of information room. At the same time, it can make the staff know better about the information room’s overall situation, and help the staff make the reasonable judgments on the main safety problems.Based on the above theoretical research, this paper gives the information room monitoring and safety evaluation platform design, from the five levels discussed the overall functional modules of the platform, the five parts of modules include platform specification interface layer, data storage, control of the core, core computing, virtual visualization module, it provides a set of data storage, data analysis, data display in one paltform for information opertion staff. Through the design and implementation of the platform we construct the virtual scene of the information room, and combined with the information room monitoring data anomaly identification results and information room fuzzy security evaluation results, the information room security operation status will be shown in the virtual scene, the information operation staff through virtual scene management of information room to make a decision for daily operation and maintenance work. |