论文部分内容阅读
分析了基于BP神经网络模型的Lyapunov指数谱计算法存在的不足,提出了一种新的基于组合策略的混沌时间序列Lyapunov指数谱计算方法.由于该方法能够同时逼近给定目标函数的非线性部分与线性部分,因而具有更高的计算精度.最后将新方法应用于Henon映射Lyapunov指数谱的计算中.通过分析与比较,表明该方法具有更高的计算精度及更强的实用性.
This paper analyzes the deficiency of Lyapunov exponent spectrum calculation based on BP neural network model and proposes a new method to calculate the Lyapunov exponents of chaotic time series based on the combination strategy.As the method can approximate the nonlinear part of a given objective function And linear part, so the new method is more suitable for the calculation of Henon mapping Lyapunov exponents spectrum.Analysis and comparison show that the new method is more accurate and more practical.