建立了Ornstein-Uhlenbeck(OU)色噪声驱动粒子的逆Kramers问题,即粒子初始处于亚稳态势阱外部,在噪声的作用下进入势阱,受左侧反射边界的影响粒子最终逃出势阱的过程.通过比较粒子在倒谐振子势场下的通过概率与粒子在亚稳态势阱中的存活概率发现:粒子在势阱中的存活概率在长时极限下趋于零且最大值比在倒谐振子势场中的通过概率的稳定值小.利用蒙特卡罗模拟计算得出粒子在势阱中的存活时间随粒子的初速度、噪声的关联时间均呈非单调变化,且存在使存活时间最长的参数,称之为"最佳初速度"以及"最佳关联时间".
Inverse Kramers problem driven by Ornstein-Uhlenbeck (OU) colored noise was built, where particle located outside of metastable well initially entered into well due to OU noise and then escaped from well under influence of left reflect boundary. After comparing pass probability in inverse harmonic potential with survival probability in metastabel well, it was found that particle survival probability tended to be zero in the long time limit and its maximum value was smaller than stable value of pass probability. Making use of the Monte Carlo method, it was calculated that particle survival time in metastable well changed non-monotonously with parameters such as particle initial velocity and noise correlation time, it was found that " best initial velocity" and "best correlation time" making survival time were the longest.