Fibonacci序列的统计属性和电子输运系数

The statistical properties and electronic transfer coefficients of Fibonacci sequence

对按膨胀规律A→AB和B→A生成的Fibonacci序列,采用一维随机行走模型数值计算了序列的自相关函数以及自行定义的准标准偏差.利用Hurst分析法研究了序列的再标度范围函数及其Hurst指数,并将结果与一维随机二元序列进行了对比.发现这些统计量有奇特的准周期振荡行为以及小于0.5的Hurst指数,直接论证了Fibonacci序列具有关联、标度不变及自相似等性质.从Anderson紧束缚模型出发,采用传输矩阵方法研究了Fibonacci序列的电子输运特性,讨论了输运系数对能量及其序列长度的依赖关系.研究表明,随着序列长度的增加具有好的透射性的电子态数量有所减少,但相对于随机序列而言,共振能态可以在更长的序列中存在,进一步证明了Fibonacci序列中存在较强的长程关联行为.

For the Fibonacci sequence constructed by following the inflation rule A→AB and B→A,using the one-dimensional random walk model and Hurst’analysis,we calculate numerically the auto-correlation function,the pseudo standard deviation of displacement defined by ourselves and the rescaled range function and investigate systematically the statistical properties. The results are compared with that of one-dimensional random binary sequence. We show that the Fibonacci sequence presents correlated behavior as well as scaling invariability and self-similarity. In addition,basing on the tight-binding model of the single electron and transfer matrix method,we study the charge transfer properties of Fibonacci sequence and discuss specially the dependence of electron transmission on energy and the length of the sequence. We find some resonant peaks can survive in relatively longer Fibonacci sequences than in random sequences,which also implies that there are long-range correlations in Fibonacci sequences.