Font Size: a A A

The Comparison Research About Plackett-Burman Data Analysis Method For Supersaturated Designs

Posted on:2009-05-08Degree:MasterType:Thesis
Country:ChinaCandidate:T MengFull Text:PDF
GTID:2120360272991479Subject:Probability theory and mathematical statistics
Abstract/Summary:PDF Full Text Request
Because of huge benefit in economic and the technical reason, it receives more and more welcome that the Analysis of Saturated Unreplicated Factorial Designs, the Plackett-Burman design has the special superiority in certain experimental situation, but at present there is not the answer that the data analysis method's comparison research about the Plackett-Burman design . This article studies the performance of two kind of Plackett-Burman design based on L12(211)and L20(219) synthetically, from which the representative eleven methods proposed in the numerical analysis method in the recent 40 years.The first chapter introduces research background, the statistical model in the research as well as the object and the goal of study preliminary.The second chapter has made the comprehensive introduction to the existing analysis method, and has made the detailed introduction to eleven method's structures and the testing procedure of the research analysis object.The third chapter has discussed the comparison research analysis method, mainly includes analysis onσvalue selection, the definition mode of active factor, the analysis on setting mode of active factor, standardized research as well as evaluation criteria.The fourth chapter studies the eleven methods to compares from the random simulations method, which uses the evaluation criteria resists errorⅠand errorⅡ, obtains a series of comparison research conclusion according to the analogue result.
Keywords/Search Tags:The Plackett-Burman design, Active factor, The standardization, Resist errorⅠ, Comparison research
PDF Full Text Request
Related items