Loading...

Economic-Statistical Design and Evaluation of Multivariate Control Charts; An Improvement of Cost Model and Constraints Approach

Ershadi, Mohammad Javad | 2009

842 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 39048 (01)
  4. University: Sharif University of Technology
  5. Department: Industrial Engineering
  6. Advisor(s): Akhavan Niaki, Taghi
  7. Abstract:
  8. Control charts are the best tools for determining the deviations in the main parameters of a process. Exponentially weighted moving average, EWMA, control charts are the best type for determining small deviations. Determining the control charts parameters by means of minimizing a cost model is economic design. Economic-statistical design is achieved by adding statistical constraints to the economic model. Average run length when the process is in control, ARL0, and average run length when the process is out of control, ARL1, are limited in the economic-statistical model. In this thesis the economic-statistical design of a multivariate EWMA control chart is considered and this model is solved by a genetic algorithm. The type of MEWMA control chart in this thesis is for detecting the deviations in the mean of the process. The Lorenzen-Vance cost function is used for estimation of the implementing of MEWMA control charts. The external costs of a company are estimated by means of a Taguchi loss function and the cost function of the model is extended by means of this approach. The parameters of the algorithm are optimized using experimental design. The method of ARL calculation in this thesis is Markov chain. The results show that the economic-statistical model leads to better statistical properties while the increases in the costs are negligible. The sensitivity analysis and some practical results are presented at the end
  9. Keywords:
  10. Exponentially Weighted Moveing Average (EWMA) ; Markov Chain ; Taguchi’s Quality Loss Function ; Genetic Algorithm ; Control Chart ; Economic Statistical Design

 Digital Object List

  • محتواي پايان نامه
  •   view

 Bookmark

No TOC