应用混沌理论,分析了网络流量,用单变量的网络流量时间序列重构与网络动力系统等距同构的相空间,进而计算了实际网络的关维数和Lyapunov指数,并证实了网络流量存在混沌特性;据此建立了基于径向基函数(RBF)神经网络的模型,并对实际网络数据流进行了预测。仿真结果表明,相对于其他前馈神经网络预测,基于混沌理论的RBF神经网络预测方法学习速度快,预测精度高。 The chaos theory is applied to analyze the network traffic. The univariate network traffic time series is used to reconstruct the isomorphic phase space of the dynamical system. Then the off-dimensional and Lyapunov exponent of the real network are calculated, and the existence of network traffic Chaos characteristics. Based on this, a model based on Radial Basis Function (RBF) neural network is established and the actual network data flow is predicted. The simulation results show that, compared with other feedforward neural network prediction, RBF neural network prediction method based on chaos theory has the advantages of fast learning speed and high prediction accuracy.