University of Twente Student Theses
Preventing Machine Crashes in a Single- and Multi-Unit System using Markov Decision Processes
Wienk, F.I. (2024) Preventing Machine Crashes in a Single- and Multi-Unit System using Markov Decision Processes.
|
PDF
4MB |
| Abstract: | In this report we formulate an optimal stopping problem for a single centrifuge that is subject to deterioration. Five different deterioration causes are identified and modelled with a discrete-time five-dimensional absorbing Markov chain. The deterioration process ends with failure, which can be a `crash' or a `run down'. Crashes are undesirable, because these can also damage neighbouring centrifuges in the system, while run downs have almost no implications. The model is extended to a system of centrifuges, in which the replacement of all non-operating centrifuges is also considered. The optimal policy for both models is proven to have a monotone structure within the partial ordering of the state and action space. The monotone structure is used in the implementation of the policy iteration algorithm to efficiently neglect certain non-optimal policies. Multiple experiments are performed to determine how the cost parameters, the number of centrifuges and the state dimensions affect the resulting policy. |
| Item Type: | Essay (Master) |
| Clients: | Urenco |
| Faculty: | EEMCS: Electrical Engineering, Mathematics and Computer Science |
| Subject: | 31 mathematics |
| Programme: | Applied Mathematics MSc (60348) |
| Link to this item: | https://purl.utwente.nl/essays/104420 |
| Export this item as: | BibTeX EndNote HTML Citation Reference Manager |
Show download statistics for this publication
Show download statistics for this publicationRepository Staff Only: item control page