A risk assessment method for power communication network

Abstract

The invention discloses a risk assessment method for a power communication network in the technical field of power communication. The present invention firstly collects the risk assessment parameters of the electric power communication network; then constructs the index database and the sample database through the risk assessment parameters; finally trains the neural network according to the sample data in the sample database, and calls the trained neural network to calculate the index database. The risk value of the power communication network corresponding to the indicator data. The present invention trains the index data through the neural network algorithm, avoiding the subjective factors introduced when the index weight is given artificially; through the network structure learning and network parameter learning, the interference of singular data is avoided, and the redundant number of hidden layer nodes is reduced , which reduces the neural network learning time and improves the network learning speed; automatically adjusts the corresponding index weights when new risk factors appear, and has good adaptability and high accuracy.

Description

technical field [0001] The invention relates to the technical field of electric power communication, in particular to a risk assessment method for an electric power communication network. Background technique [0002] The power communication network is related to every household, and its security is very important. As the scale of power communication network continues to increase, the topology structure becomes more and more complex, and the uncertainty of network operation also increases. Scientifically evaluate the risks existing in the power communication network, and use the evaluation results as the basis for selecting safety precautions for the power communication network. By improving management measures, risks can be effectively reduced or avoided, and system security can be improved. Stable operation has important practical significance. This patent proposes a risk assessment method for a power communication network. [0003] Risk assessment is mainly to comprehe...

AI Technical Summary

Problems solved by technology

However, this invention belongs to the field of information security. For the risks caused by network attacks, it does not consider the operating status of power communication network equipment and risk factors in the environment.

[0007] Risk assessment of optical fiber protection channel based on PCA-RBF neural network (Journal of North China Electric Power University, 2007, 34[5]: 90-94): Combining principal component analysis method PCA with RBF neural network, using PCA to approximate the number of indicators Jane, removed multiple indicators, greatly reduced the dimension of the RBF input vector, and improved the efficiency of the method. However, the comprehensiveness of the indicator system has declined, and the accuracy of the evaluation results will also be affected.

[0008] In terms of evaluation methods, there are also some problems in the above-mentioned existing risk assessment methods. Among them, the risk assessment method based on expert experience needs to artificially give the weight of each indicator, so it is impossible to avoid the interference of human subjective factors, and the credibility of the evaluation results cannot be avoided. Its bottleneck is that the evaluation results are subjectively arbitrary due to the subjective weighting value of experts. This shortcoming cannot be fundamentally improved by taking measures such as increasing the number of experts and carefully selecting experts; while the risk assessment method based on statistical data although Based on the statistical data of the indicators, the objectivity of the evaluation results is guaranteed, but due to the different emphases, this method can not avoid the disadvantage of singular data interfering with the evaluation results while highlighting the advantages of objectivity, and sometimes the evaluation results will be contrary to the actual situation; Patent 1 and Patent 2 are mainly aimed at the risks of computer software and network attacks, and both do not consider the risk factors in the power communication network equipment failure and network operating environment; although the method proposed in Document 3 greatly reduces the dimensionality of the RBF input vector increased the efficiency of the method, but the comprehensiveness of the indicator system has declined

Images