Parts | Days | Selection | Search | Downloads | Help
AKSOE: Arbeitskreis Physik sozio-ökonomischer Systeme
AKSOE 10: Social-, Information-, and Production Networks
AKSOE 10.4: Talk
Wednesday, March 28, 2007, 17:45–18:15, H8
Robustness and Dynamical Behaviour of Small-Scale Manufacturing Networks — •Reik Donner1, Uwe Hinrichs2, and Bernd Scholz-Reiter2 — 1Institute of Transport and Economics, Technical University Dresden, Andreas-Schubert-Str. 23, 01062 Dresden — 2BIBA, University of Bremen, Hochschulring 20, 28359 Bremen, Germany
Intrinsic instabilities of production networks are a serious economic danger especially for small companies. On the one hand, random fluctuations of processing times cause an immediate instability which can lead to production breakdowns due to a lack of material and result in an intermittent accumulation of orders. On the other hand, event-discrete simulations indicate that such fluctuations may also have a constructive effect on the dynamics: If a push strategy is applied for the production process, long down times of the machines after each processed workpiece may lead to a successive increase of the stocks of semi-finished material. This increase is found to be decelerated by fluctuating processing times. In the more realistic case of a pull control, improper lot sizes result in an aperiodic behaviour even for constant processing times. The critical parameters associated with the different effects are identified for two different network topologies: linear supply chains and symmetrically interacting manufacturers. For a quantitative characterization of the dynamics, we apply sophisticated nonlinear methods of time series analysis, such as concepts basing on a symbolic dynamics description as well as recurrence quantification analysis.