|
Type of Document Thesis Author Sciullo, Charles Augustine URN etd-11132007-155143 Title Augmentation of Second Order Designs for Validation and Refinement Degree Master of Science Department Industrial and Manufacturing Engineering, Department of Advisory Committee
Advisor Name Title James R. Simpson Committee Chair Amy Chan Hilton Committee Member Joseph J. Pignatiello, Jr. Committee Member Keywords
- Augmentation
- Second Order Desgins
- Validation
- Refinement
Date of Defense 2007-10-18 Availability unrestricted Abstract Validation is an important step that is often left out of modeling due to lack of time, lack of funding, or just plain ignorance. Many industries that practice design of experiments leave out this important step because the benefits are not visible for the model that is often created. In addition, there appears to be no clear way to measure whether or not the validation points truly validate the model. This paper focuses on a new way of validation that will both refine the model, yielding more information, and validate the model. This new approach of validation is a simple way of improving the model and making sure that it does what it was designed to do. This validation procedure is only used for 2nd order models that are based off of a central composite design, but have the possibility of picking up 3rd order terms that the model was unable to identify without the extra validation points. Validation points are chosen through a genetic algorithm using space-filling and orthogonality criteria. A new set of points that is different from that of the design is derived. These points are better than just randomly selecting points that have no meaning. Validation is a necessary step to eliminate many risks. The procedure defined here is useful for determining 3rd order terms and their validity in a model.Files
Filename Size Approximate Download Time (Hours:Minutes:Seconds)
28.8 Modem 56K Modem ISDN (64 Kb) ISDN (128 Kb) Higher-speed Access SciulloCThesis.pdf 916.84 Kb 00:04:14 00:02:10 00:01:54 00:00:57 00:00:04