Machine-learning-based approach for parameterizing material flow simulation models
| Autor: | Ludwig Trauner, Kilian Vernickel, Georg Hoellthaler, Lukas Bank, Giuseppe Sansivieri, Christian Härdtlein, Julia Berg, Laura Brunner, Jan Fischer |
|---|---|
| Přispěvatelé: | Publica |
| Jazyk: | angličtina |
| Rok vydání: | 2020 |
| Předmět: |
Process (engineering)
Computer science business.industry Simulation modeling maschinelles Lernen Industrie 4.0 Machine learning computer.software_genre Material flow Set (abstract data type) General Earth and Planetary Sciences Production (economics) Artificial intelligence Discrete event simulation business computer Simulation General Environmental Science |
| Popis: | Discrete Event Simulation (DES) provides a far-reaching set of methods for planning and improving structures and processes of real factories. However, the use of DES is hampered by the high effort required for creating and parameterizing the corresponding models. Regarding the material flow of production systems with different products and small lot sizes, the expenditure for identifying current and historical processes and their parameters in the available data is time-consuming. Therefore, this paper presents an approach where machine learning algorithms identify these process specific parameters. Following that, the algorithms are integrated into the material flow simulation to parameterize the elements in the model. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|