Security Risk Evaluation Of Electric Power Telecommunication Networks Based On Neural Network

Security risk evaluation of electric power telecommunication networks is important for decision-making of it.In this paper a security risk evaluation method based on radial basic function (RBF) neural network is presented. It can quantificationally express the important degree of the index architecture and accurately calculate the risk result of the electric power telecommunication networks, and effectively enhance the counting efficiency. At the same time, in view of the influence factor and the sample data characteristic,these methods which are algorithm of filling missing values based on rough set,standardization formulas about dimension and algorithm of repeatedly using the evaluation samples have been proposed. In view of establishment of the evaluation model, the information entropy and principal component analysis is used to ascertain the number of the input nodes, dynamic clustering is used to ascertain the number and center of the hidden nodes, the evaluation request of the power communications network is used to ascertain the number of the output nodes.Finally, we carry on the training and test to the RBF neural network, and security risk evaluation to the electric power communication networks. The result indicates that the method is feasible and the superiority in evaluation precision and the adaptive ability.