关键词: 相依网络, 预测, 演化过程, 李雅普诺夫指数, 长短期记忆人工神经网络

Abstract: Dynamic and accurate operation situation prediction of control system is the key basis for collaborative decision-making among relevant units of air transport system. In this paper, a control-aircraft state interdependent network model was established based on the conflict relation of aircrafts, the command relation of controllers to aircrafts and the transfer relation of controllers. The evolution law of the interdependent network is explored and analyzed. Correlation analysis and principal component analysis are used to prove the rationality of the five indicators selected It is proved that each time series has chaotic characteristics by calculating the maximum Lyapunov exponent of topological indexes such as average node degree and average point weight. And the long short-term memory method is selected to predict the evolution law of each time series. The simulation results show that the proposed method can effectively predict the evolution process of the control system. The prediction accuracy of the five indicators has a probability of 96% within 20%.

Key words: interdependent network, prediction, evolution process, Lyapunov exponent, long short-term memory