发表机构
Institute for Nuclear Research（核物理研究所）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本文推广祖巴列夫非平衡统计算子方法，构建三参数模型统一处理首次通过时间、停留时间和绝对极值等边界泛函，并利用杜布h变换实现非马尔可夫输运，用于优化周期性势中的随机粒子分离。
AI 中文摘要
本文发展了祖巴列夫非平衡统计算子方法的一种推广，用于同时包含轨迹的加性和非局部边界泛函的情形。采用统一的热力学方法，构建了一个非平衡系统的三参数模型，其中包括首次通过时间、在给定水平以上的停留时间以及过程的绝对极值。证明了轨迹的最大信息熵方法与唐斯克-瓦拉丹大偏差形式体系之间的对应关系。利用杜布h变换，证明了固定历史的极值泛函会在系统中引发有效的非马尔可夫输运，并具有对历史记录的动态适应性。讨论了所发展装置在“大时间”域中的适用性标准，以及其在周期性势中优化随机粒子分离的可能性。
英文摘要
This paper develops a generalization of Zubarev's nonequilibrium statistical operator method for the case of simultaneous inclusion of additive and nonlocal boundary functionals of trajectories. Using a unified thermodynamic approach, a three-parameter model of nonequilibrium systems is constructed, including the first-passage time, the dwell time above a given level, and the absolute extremum of the process. A correspondence is demonstrated between the maximum information entropy method for trajectories and the Donsker-Varadan large deviation formalism. Using the Doob`s h-transform, it is demonstrated that fixing extremal functionals of history induces efficient non-Markovian transport in the system with dynamic adaptation to historical records. Criteria for the applicability of the developed apparatus in the "large time" domain and the possibilities of its use for optimizing stochastic particle separation in periodic potentials are discussed.
Comments33 pages, 5 figures
