Production lines with unreliable machines have received a great amount of attention
in the literature. Especially, two-station systems have mostly been studied because
such systems are easier to handle when compared to the longer lines. In literature,
longer lines are usually evaluated by a decomposition algorithm, whereby the long
line is partitioned into chunks of two-station lines. Decomposition algorithms require
intermediate buffer storages of capacity at least two or three. The trends in modern
manufacturing practices, on the other hand, such as the Toyota Production System,
dictate that intermediate storages be eliminated. Our work studies multi-station lines
with no intermediate storage. We develop software to automate the generation of
transition probability matrices to allow the analysis of system behavior.
The algorithm allows the use of software packages to handle computations and to
solve for exact solutions. Long-run behavior is obtained via the algorithm developed
in the computational environment MATLAB. The purpose is to analyze the system
performance measures such as starvation and blockage times of stations, production
rate and work-in-process.
In addition, the production rate and the work-in-process measures over failure and
repair probabilities are curve-fit to establish simple and useful empirical formulas for
lines consisting three, four and five identical stations. Numerical analyses show that
the proposed algorithm is effective for exact solutions and the suggested formulas are
valid for approximate solutions.
| Identifer | oai:union.ndltd.org:METU/oai:etd.lib.metu.edu.tr:http://etd.lib.metu.edu.tr/upload/12612227/index.pdf |
| Date | 01 July 2010 |
| Creators | Cetinay, Hande |
| Contributors | Kayaligil, Sinan |
| Publisher | METU |
| Source Sets | Middle East Technical Univ. |
| Language | English |
| Detected Language | English |
| Type | M.S. Thesis |
| Format | text/pdf |
| Rights | To liberate the content for public access |
Page generated in 0.2124 seconds