Vinsamlegast notið þetta auðkenni þegar þið vitnið til verksins eða tengið í það: http://hdl.handle.net/1946/16735
The topic of this thesis is an analysis of Alcoa Fjarðaál´s production process, from tapping to casting, emphasizing on effects of production planning. An optimization model for cell batching was developed and its results compared to the current algorithm used. The model was also used to estimate impact of improved production planning in the cast house. The model developed is a binary integer model, a so called Set Partitioning Problem. It was solved by use of the Gurobi solver.
Currently Alcoa Fjarðaál is using an optimization algorithm developed by Alcoa and is a part of their ERP system and one of this thesis’ objective is to estimate the quality of its solutions. A comparison of tapping plans from the two models showed that the current model delivers much better results than a naive algorithm that batches 3 adjective cells together but does not deliver an optimal solution.
Impact of errors in material requirement planning in cast house on the tapping plan suggested by the model was estimated by generating tapping plans based on the actual usage and assumptions used by Alcoa´s employees. It showed that errors in material requirement planning have a significant impact on the tapping plan generated.
Potential gain of improving allocation of metal to casting lines in the furnace area of the cast house was estimated. It revealed that improved planning of the tapping process can have a significant impact.
|Improvement Opportunities in Aluminium Production-1.pdf||2.76 MB||Lokaður til...10.10.2033||Heildartexti|