Dresden 2020 – wissenschaftliches Programm
Die DPG-Frühjahrstagung in Dresden musste abgesagt werden! Lesen Sie mehr ...
DY 42.3: Vortrag
Mittwoch, 18. März 2020, 15:45–16:00, ZEU 160
Memory capacity of a flow network — •Komal Bhattacharyya1, David Zwicker1, and Karen Alim1,2 — 1Max Planck Institute for Dynamics and Self-Organization, 37077 Göttingen, Germany — 2Physik-Department, Technische Universität München, Garching, Germany
The slime mould Physarum polycephalum is a very simple unicellular but seemingly intelligent organism with a network-like body. Its complex behaviour requires the ability to propagate, store and process information. Recently, it has been shown that Physarum propagates information about stimuli with the fluid flows throughout its network. Most inspiringly, Physarum was observed to adapt its networks tube radii globally in response to stimuli, reaching a steady-state as a long term response that keeps a memory of the stimuli in its network morphology. Inspired by this observation we here investigate the capacity to store information about previous stimuli in the morphology of an adaptive flow network. We model the organism as a flow network whose radii can change when optimising the network to have the least energy dissipation. We observe how the system reacts to localised changes and the timescale of its responses to applied stimuli by numerical simulation. Through theoretical understanding, we aim to pinpoint the information storing and processing capabilities of adaptive flow networks in general and Physarum networks specifically.