h1

TY  - THES
AU  - Stehr, Ernst-August
TI  - Nutzung der ereignisdiskreten Simulation für die operative Produktionsplanung in der mehrstufigen Einzel- und Kleinserienfertigung
PB  - Rheinisch-Westfälische Technische Hochschule Aachen
VL  - Dissertation
CY  - Aachen
M1  - RWTH-2023-03051
SP  - 1 Online-Ressource : Illustrationen, Diagramme
PY  - 2022
N1  - Veröffentlicht auf dem Publikationsserver der RWTH Aachen University 2023
N1  - Dissertation, Rheinisch-Westfälische Technische Hochschule Aachen, 2022
AB  - Manufacturing companies face stiff competition in global markets. One way of differentiating themselves is by fast and reliable delivery of goods. Studies confirm the paramount importance of these in the manufacturing control departments of the industry. These objectives are either accomplished by technological, processual or organisational changes. The paced assembly lines of the automotive industry and its successors of Lean production and mixed model assembly have achieved tremendous successes in supplying the world with its goods. However, these methods cannot easily be transferred to a job shop with single products or small series as many very different products have to share the same factory. An alternative approach is the usage of software for production planning. This started with ERP (transactional and accounting) via MRP (material planning without capacities) to APS (simultaneous planning of all relevant factors). The last option however reacts very nervous to minor production deviations, leading to massive changes in the forecasted delivery times. Discrete Event Simulation (DES) has been used since the 1960’s to plan factories under high levels of uncertainty. This thesis extends this basic algorithm to allow the operative planning of multi-level job-shops. Therefore, simultaneously personal-, machine- and material availability has to be considered. The developed algorithms is for the first time able to coordinate order networks of multiple components for one assembly instead of just reacting to the manufacturing progress of the parts. The resulting algorithm was benchmarked against a version of a typical APS algorithm. The latter was re-implemented and tested against the new DES for the same real-world manufacturing data set. Results indicate that DES is slightly better in terms of utilization, throughput and inventory (5-10
LB  - PUB:(DE-HGF)11
DO  - DOI:10.18154/RWTH-2023-03051
UR  - https://publications.rwth-aachen.de/record/954251
ER  -