2007/12/21 by A. E. Allahverdyan, David B. Saakian, D. B. Saakian
Engineering · Mathematics · Neuroscience · Physics and Astronomy · #Adiabatic process #Advanced Thermodynamics and Statistical Mechanics #Artificial intelligence #Classical mechanics #Computer science #Control (management) #Control engineering #Control theory (sociology) #Engineering #Entropy (arrow of time) #Entropy production #Ergodic theory #Feedback control #Hamiltonian (control theory) #Hamiltonian system #Mathematical analysis #Mathematical optimization #Mathematics #Negentropy #Neural dynamics and brain function #Physics #Statistical physics #Thermodynamics #cond-mat.stat-mech #stochastic dynamics and bifurcation
paper · pdf · doi:10.1209/0295-5075/81/30003
published as Europhys. Lett. 81, 30003 (2008) · 6 pages, 1 figure
openalex publication_date 2007/12/21 · arxiv created 2008/11/15 · arxiv updated 2009/12/01 · openalex created_date 2025/10/10 · openalex updated_date 2026/08/06
We study adaptive control of classical ergodic Hamiltonian systems, where the controlling parameter varies slowly in time and is influenced by the system's state (feedback). An effective adiabatic description is obtained for slow variables of the system. A general limit on the feedback-induced negative-entropy production is uncovered. It relates the quickest negentropy production to fluctuations of the control Hamiltonian. The method deals efficiently with the entropy-information trade-off.