Joint Maintenance Interval and Spare Parts Optimization using a Discrete-Event Simulation Model
MetadataShow full item record
The goal of this report is to use discrete-event simulation (DES) as a method for optimizing maintenance strategies, such as spare parts levels and maintenance intervals. Firstly, the au- thor argues for spare parts optimization with a DES in Visual Basic for Applications (VBA). The models and assumptions that are needed for developing such a model are explained. Further- more, this report elaborates on how a DES can be coded in VBA. Lastly, several methods for optimizing both speed and decision variables of a DES are introduced.The report shows how a DES can be coded and which models and assumptions can be used in developing such a simulation. A specific focus is on the design of the pending-event set (PES), which is the core of the DES. Several different designs are tested in different situations in order to determine their performance. The results show that the performance of these methods vary in each situation, and therefore the designer of a DES should determine the characteristics of the DES, before an appropriate PES method can be chosen. This thesis shows that a simplified genetic algorithm can be used in order to find good results in a faster and more structured way than a trial-and-error method. It furthermore shows that this genetic algorithm can be used for joint optimization of preventive maintenance interval, the overhaul interval, spare order thresh- old and stock levels.The author concludes the report with recommendations for further work. On the practical side, the impact of different PM strategies on stock levels should be researched. Furthermore, the research to including condition-based maintenance (CBM) in a model like this should be taken a step further with a more complex model for CBM. On the theoretical side, the PES methods should be more thoroughly studied. More functions to manipulate the PES and the required memory space should be included in further research. Lastly, the author believes that the simplified genetic algorithm can be further improved, which can be a focus topic in further research.