We investigate thermodynamics of feedback processes driven by measurement. Regarding system and memory device as a composite system, mutual information as a measure of correlation between the two constituents contributes to the entropy of the composite system, which makes the generalized total entropy of the joint system and reservoir satisfy the second law of thermodynamics. We investigate the thermodynamics of the Szilard engine for an intermediate period before the completion of cycle. We show the second law to hold resolving the paradox of Maxwell's demon independent of the period taken into account. We also investigate a feedback process to confine a particle excessively within a trap, which is operated by repetitions of feedback in a finite time interval. We derive the stability condition for multi-step feedback and find the condition for confinement below thermal fluctuation in the absence of feedback. The results are found to depend on interval between feedback steps and intensity of feedback protocol, which are expected to be important parameters in real experiments.
Markov vs. nonMarkovian processes A comment on the paper Stochastic feedback, nonlinear families of Markov processes, and nonlinear Fokker-Planck equations by T.D. Frank
本文旨在纠正T. D. Frank.我们的评论集中在一个非线性马尔可夫过程和一个非线性福克-普朗克方程的索赔。首先,转移密度中的记忆被误认为是马尔可夫过程。第二,弗兰克假设人们可以从查普曼-柯尔莫哥洛夫方程推导出福克-普朗克方程,但没有证据表明查普曼-柯尔莫哥洛夫方程存在于记忆依赖过程中。基于1-点概率密度的非线性扩散偏微分方程,提出了一个非线性马尔可夫过程。我们表明,无论哪一个初始值问题可以解决的1点密度,所产生的随机过程,必然定义的转移概率,要么是一个普通的线性生成的马尔可夫过程,否则是一个线性生成的nonMarkovian过程的记忆。我们提供了明确的例子,扩散系数,反映了马尔可夫和记忆依赖的情况下。因此,跃迁密度既不存在非线性马尔可夫过程,也不存在非线性Fokker-Planck方程。文献中的混乱部分源于将1点概率密度的非线性扩散方程标记为非线性福克-普朗克,而1点密度和1点密度的运动方程都没有定义随机过程,并且Borland错误地将来自非线性扩散方程的平移不变1点密度识别为条件概率密度。在附录中我们推导了Fokker-Planck偏微分方程和Chapman-Kolmogorov方程。有限记忆随机过程