基于组合策略的Lyapunov指数谱的计算

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.

Calculation of Lyapunov Exponents Based on Combination Strategy

The defect of the Lyapunov exponents calculation method based on BP neural network model is analyzed, and a novel method to calculate Lyapunov exponents of chaotic time series based on combination strategy is proposed. Because this method can approximate the nonlinear part and linear part of a given objective function at the same time, its calculation preciseness is higher than the method based on BP neural network. In the end, the new method is applied to computing the Lyapunov exponents of Henon map. By means of analyzing and comparing, it shows that the suggested method is more precise and practicable in the calculation of Lyapunov exponents.