Dresden 2020 – wissenschaftliches Programm

Die DPG-Frühjahrstagung in Dresden musste abgesagt werden! Lesen Sie mehr ...

Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe

DY: Fachverband Dynamik und Statistische Physik

DY 36: Stochastic Thermodynamics

DY 36.6: Vortrag

Mittwoch, 18. März 2020, 11:15–11:30, ZEU 147

Non-reciprocal hidden degrees of freedom: an attempt to define entropy production in non-Markovian systems — •Sarah A. M. Loos, Simon M. Hermann, and Sabine H. L. Klapp — TU Berlin, Hardenbergstr. 36, 10623 Berlin

Real-world stochastic systems are often non-Markovian. This might be due to hydrodynamic backcoupling, viscoelastic effects, persistence in active swimmers, or an external feedback control loop acting on the system. Despite the omnipresence of memory, the incorporation of non-Markovian dynamics in the framework of stochastic thermodynamics is poorly understood [1,2]. In fact, fundamental problems remain, which are associated with the acausality of the backward process in the total entropy production functional. We discuss this crucial issue focusing on the case of discrete time delay, and show the implications of different approaches. As a first example, we review the outcome of a direct calculation on the basis of the acausal path integrals, which requires redefining the definition of entropy production, and, in fact, yields a functional which by construction cannot be calculated for any nonlinear system. Furthermore, we suggest a Markovian embedding approach [3]. While this strategy allows us to employ the standard formulae and is technically much simpler, it demands the interpretation of entropy production of auxiliary variables. For the case of a feedback controller, we offer an appropriate interpretation.

[1] Munakata, Rosinberg, PRL 112, 180601 (2014).

[2] Loos, Klapp, Sci. Rep. 9, 2491 (2019).

[3] Loos, Hermann, Klapp, preprint: arXiv:1910.08371 (2019).