|
|
Type of Document Thesis Author Guo, Yong Author's Email Address guo@eng.fsu.edu URN etd-11212003-230640 Title Construction of Efficient Fractional Factorial Mixed-Level Designs Degree Master of Science Department Industrial and Manufacturing Engineering, Department of Advisory Committee
Advisor Name Title James R. Simpson Committee Chair Joseph J. Pignatiello, Jr. Committee Member Samuel A. Awoniyi Committee Member Keywords
- Design Of Experiments
- Genetic Algorithm
- Efficient Mixed-Level Designs
- Balance Coefficient
Date of Defense 2003-11-21 Availability unrestricted Abstract Mixed-level factorial designs are experimental designs whose factors have different numbers of levels. These designs are very useful in experiments involving both qualitative and quantitative factors. One design approach is to run all possible combinations of the factor levels. However, as the number of factors or factor levels increases, the number of experiments increases dramatically. As a result, research has focused on developing orthogonal or near-orthogonal fractional factorial designs. The property of design balance, that the same number of runs is performed for each factor level, has been maintained in currently proposed designs. In some cases, maintaining balance requires too many experimental runs. The objective of this thesis is to develop fractional mixed-level factorial designs with economical run size that have desirable properties associated with near-balance and near-orthogonality. Two criteria are developed to assess the degree of near-balance for comparing and constructing designs. A modified J2-optimality criterion is used for comparing design near-orthogonality. These criteria are combined to assess different design alternatives. A genetic algorithm is then used to build designs with the most desirable combination of near-balance and near-orthogonality.Files
Filename Size Approximate Download Time (Hours:Minutes:Seconds)
28.8 Modem 56K Modem ISDN (64 Kb) ISDN (128 Kb) Higher-speed Access YongGuo_Thesis.pdf 587.84 Kb 00:02:43 00:01:23 00:01:13 00:00:36 00:00:03