- 期刊
Industrial Internet Security Situation Prediction Based on NDPSO-IAFSA-LSTM
摘要
The traditional prediction method takes a long time to train, and the prediction accuracy is not high. Because of the above problems, to further improve the accuracy and efficiency of Industrial Internet security situation prediction, this paper improves the situation prediction method of a Long Short-Term Memory network(LSTM) based on the neural network, which can better process the data of time series. It proposes an Industrial Internet security situation prediction method based on NDPSO-IAFSA-LSTM by integrating nonlinear dynamic particle swarm optimization, improved artificial fish swarm algorithm, and LSTM. Furthermore, NDPSO and IAFSA are used to solve the problem that the two algorithms easily fall into local extremum during optimization and optimize the parameters of the LSTM network. Finally, this paper verifies the designed Industrial Internet security situation prediction method in a simulated network environment. The experimental results show that this method’s mean square error and absolute error are 0.005 and 0.0209, respectively, which are less than the error of NDPSO-LSTM, LSTM, and IAFSA-LSTM prediction methods, and improve the accuracy of the Industrial Internet security situation prediction method.