13:30 Uhr
Automated Generation and Simulation of Hyper Heuristics for Stochastic Multi-Mode Multi-Project Resource-Constrained Project Scheduling Problems with Setup Times
Rico Zöllner | Technische Universität Dresden
Details anzeigen
Autor:innen:
Rico Zöllner | Technische Universität Dresden
Mathias Kühn | Technische Universität Dresden
Konrad Handrich | Technische Universität Dresden
Thorsten Schmidt | Technische Universität Dresden
A simulation framework is presented which covers both generation and simulation of production planning and control problems which include transfer times and stochastic influences and therefore extend classical multi-mode multi-project RCPSPs. This allows for systematic and in-depth investigations of the quality and the behaviour of heuristics. In addition, the automated design of heuristics based on Boolean operators applied to relations of problem specific quantities leads on average to better results than a manual selection and adjustment of heuristic strategies.
Manuscript herunterladen
14:00 Uhr
Iterative optimierungsbasierte Simulation in der Praxis – Simulation der Oberflächenveredelung der Salzgitter Flachstahl GmbH
Patrick Oetjegerdes | TU Braunschweig
Details anzeigen
Autor:innen:
Patrick Oetjegerdes | TU Braunschweig
Christian Weckenborg | TU Braunschweig
Thomas S. Spengler | TU Braunschweig
Udo Vogeler | Salzgitter Flachstahl GmbH
Mike Molzberger | Salzgitter Flachstahl GmbH
The Salzgitter Flachstahl GmbH plans a shift in their production program, which might require adaptions in the logistics. A discrete event simulation model is used to evaluate the impact of the new production program on logistics and the effects of various adaptations. In practice, however, optimization algorithms are applied in this system to facilitate resource planning, e.g., for loading train wagons or crane scheduling. To develop a valid production system model, it was necessary to implement optimization algorithms directly within the simulation and execute them iteratively during the simulation run. Using this approach, different adaptions of the logistic system could be evaluated. The approach to implementing iterative optimization-based simulation in practice and the lessons learned are reported.
Manuscript herunterladen
14:30 Uhr
DISPO 4.0 - Simulationsbasierte Optimierung von Bestelllosgrößen
Alexander Schmid | Fraunhofer Austria Research GmbH
Details anzeigen
Autor:innen:
Alexander Schmid | Fraunhofer Austria Research GmbH
Alexander Granig | Technische Universität Wien
Thomas Sobottka | Fraunhofer Austria Research GmbH
Martin Riester | Fraunhofer Austria Research GmbH
Wilfried Sihn | Fraunhofer Austria Research GmbH
This paper presents the development and evaluation of a digital planning method for the operational purchasing that enables an automated and product-specific optimized material requirements planning. With materials planning today facing a volatile and complex market environment, the trend towards automated processes in procurement is the key to competitiveness. Although simulation-based optimization for one of the key tasks in material requirements planning, purchase order lot size planning, offers great potential, these digital tools are not yet readily available and thus not widely used in the industry. Thus, this paper develops a simulation-based optimization of purchase orders and evaluates the method in two case-studies in the capital good industry, using 15 purchase order lot sizing algorithms. The results point to a significant cost savings potential and provide insights into the comparative benefits of the algorithms.
Manuscript herunterladen