Network Fault Diagnosis Based on Improved BP Neural Network
Ping Chen · Computer and Digital Engineering · 2012
To solve the problem that network fault diagnosis rules are hard to get,a fault diagnosis method based on improved BP neural network is presented.The fault information is taken as samples to train the BP neural network,based on that,the relations between the network fault information and fault patterns are established utilizing the self-adaptation and nonlinearity mapping functions of the neural network.At the same time,Levenberg-Marquardt algorithm is applied to avoid that the learning algorithm of BP neural gets into local minimum,and to improve the efficiency and accuracy of the network.In addition,the thought of early-stop is adopted to improve the generation of the BP neural network.Finally,a practical example indicates the feasibility and validity of the method.