含噪声混沌信号最大Lyapunov指数计算

Computation of the largest Lyapunov exponent in noisy chaotic signal

提出一种计算含噪声混沌序列最大Lyapunov指数的改进Rosentein算法,通过将原始含噪声的相空间与局部投影去噪后的相空间合并,使邻近点的选择更为准确,减小噪声的影响提高计算精度.分别对加入白噪声的Henon混沌序列和逐月太阳黑子混沌序列进行仿真,证明此方法有效性.

An improved Rosentein algorithm is presented for calculating the largest Lyapunov exponent of the noisy chaotic time series. Through combining the noisy phase space and the phase space of the local geometric projection for noise reduction, this method selects more accurate neighborhood points, reduces the noise influence and increases the computation accuracy. The validity of this method is shown by simulations results of its application to chaotic time series generated by Henon map with white noise and the observed monthly sunspot time series.