Keywords and Phrases
"Preventive maintenance is an important tool that increases the reliability of the production system by reducing downtime due to failures. In the literature, maintenance and replacement policies for productions systems have been widely studied and modeled. Traditionally preventive maintenance has focused only on minimizing expected costs without considering variability in costs. Cost variability is also commonly known as risk. In a 2003 paper, Chen and Jin used variance of costs as a measure of risk for formulating preventive maintenance policies. The variance criterion that they used ignored the probability of costs exceeding monthly or yearly budgets provided to managers. The goal of the present work is to develop a performance metric for preventive maintenance that will not only consider long-run average cost but also minimize the chance that costs will exceed pre-specified budgets. Therefore the model introduced here uses a relatively less known risk metric called semivariance. The semivariance model developed here relies on an objective function that combines average cost with risk via the framework developed by Markowitz in 1952. It uses renewal theory and semi-Markov decision processes to develop mathematical expressions for the average cost and risk. These mathematical models are implemented within MATLAB, but they can also be implemented in spreadsheet software such as Microsoft Excel. We show via numerical experiments that the semivariance-penalized model outperforms cost-based and variance penalized models"--Abstract, page iii.
Long, Suzanna, 1961-
Murray, Susan L.
Engineering Management and Systems Engineering
M.S. in Engineering Management
Missouri University of Science and Technology
ix, 47 pages
© 2011 Venkata Manojramam Tirumalasetty, All rights reserved.
Thesis - Open Access
Maintenance -- Mathematical models
Service life (Engineering) -- Forecasting
Print OCLC #
Electronic OCLC #
Link to Catalog Record
Tirumalasetty, Venkata Manojramam, "Risk-sensitive preventive maintenance policies using semivariance" (2011). Masters Theses. 6921.