一种新的混沌序列随机性刻画及提高方法

通过定义混沌序列中元素之间的邻接关系,提出了一种基于网络邻接图的方法来刻画此序列的随机性.通过数值模拟发现此方法具有很好的有效性,且与传统的Lyapunov指数判别法相一致,尤其在Lyapunov指数难以求得的情况下,此方法具有很好的参照性.通过考察网络邻接图的度分布、聚类系数和平均路径长度,发现混沌序列的网络邻接图是无标度网络,并且具有明显的小世界特性.利用网络邻接图的无标度性,适当减少网络邻接图中为数不多的度大的点可以有效提高混沌序列的随机性.

A New Method to Characterize and Enhance Randomness of Chaotic Sequences

By defining adjacent relations among the components of a chaotic sequence,we introduce a method based on network adjacent graph to depict the randomness of this sequence.Through numerical simulations we find that this method has good effectiveness,and is consonant with the traditional Lyapunov exponent method,particularly for the case when the Lyapunov exponents are difficult to calculate.By examining the degree distribution,clustering coefficient and average path length of network adjacent graph,we find th...