随机双指数记忆耗散系统的非马尔可夫扩散

针对具有双指数耗散记忆核函数的两自由度耦合系统, 本文利用Laplace变换导出了热宽带噪声激励下该系统响应二阶矩的解析表达式. 并观察到位移二阶矩不同于单自由度情形下的反常扩散:<x²(t)> ∝ t^α (0<α<2, α≠1), 而是随时间及噪声等参数变化呈现普遍的振荡扩散现象.分析可得, 阻尼耦合因子B使粒子远离简谐势场的束缚, <x²(t)>随B的增大扩散加剧而摩擦系数增大却使其趋于平稳状态.进一步, 若两热噪声互关联时, 较小的互关联时间对二阶矩的影响较大, 反之作用较小. 伴随互关联强度递增, 位移二阶矩的扩散加剧, 位移间的相关性加强, 与物理直观相符. 关键词： 热噪声 非马尔可夫扩散 广义朗之万方程 关联性

Non-Markovian diffusion of the stochastic system with a biexponentical dissipative memory kernel

In this paper, second-moments of the responses are analytically solved by the Laplace transform in a coupling two-degree-of-freedom system with a biexponentical dissipative memory kernel function driven by a thermal broadband noise. The mean square displacement <x²(t)> is different from anomalous diffusion (i.e. <x²(t)> ∝ t^α (0<α<2, α≠1)), which is produced by the single-degree-of-freedom generalized Langevin equation. The oscillation-diffusion of <x²(t)> with the change of time and noise parameters is observed generally. According to our analysis, a particle confined by the harmonic potential can escape with the help of the coupling-damping factor B. The diffusion of <x²(t)> aggravates with B increasing. However, <x²(t)> tends to the stationary state with the increase of the friction coefficient Further, if the two thermal noises are in cross-correlation, smaller cross-correlation time has a deeper influence on second-moments. Meanwhile, the diffusion aggravates and the cross-correlation between two displacements strengthens markedly with cross-correlation strength increasing. It is consistent with physical intuition.
Keywords： thermal noise non-Markovian diffusion generalized Langevin equation correlation