Font Size: a A A

Two High-level Factors Are Designed In Different Regions With A Clean Effect

Posted on:2017-04-29Degree:MasterType:Thesis
Country:ChinaCandidate:X F TengFull Text:PDF
GTID:2359330485974440Subject:Statistics
Abstract/Summary:PDF Full Text Request
Fractional factorial (FF) designs are widely used in the design of experiments. When the level of some of the factors in the experiment is difficult to control or change, it is almost impossible to carry out complete randomization in the experiment. Then we will give up complete randomization, and use fractional factorial split-plot design. In the actual experiment, there are often two-level and four-level even higher-level factors at the same time, then the mixed-level split-plot design can be used. The mixed-level design can be obtained by the method of replacement from two-level design.This article investigates the mixed-level fractional factorial split-plot designs with clear effects that have a four-level factor and some two-level factors in the whole-plot, an eight-level factor and some two-level factors in the sub-plot, denoted as 2(n1+n2)-(k1-k2)4w18s1.We give the conditions for the design that contains different clear effects and the construction method of the designs. It is divided into five chapters.Chapter 1 has four sections. Section 1.1 introduces the definition of factorial design. Section 1.2 introduces the two-level factorial design and some criteria for se-lecting good designs. Section 1.3 introduces mixed-level factorial design. Section 1.4 introduces the properties and construction method of the split-plot design.Chapter 2 introduces the properties a,nd construction method of the mixed-level fractional factorial split-plot design.Chapter 3 investigates the conditions for resolution ? 2?(n1+n2)-(k1+k2)4w18s1 design containing different clear effects and the construction method of 2?(n1+n2)-(k1+k2)4w18s1 design that contains all kinds of clear effectsChapter 4 investigates the conditions for resolution ? 2?(n1+n2)-(k1+k2)4w18s1 design containing different clear effects and the construction method of 2?(n1+n2)-(k1+k2)4w18s1 design that contains all kinds of clear effects.Chapter 5 gives a brief summary of the whole article.
Keywords/Search Tags:Clear, Mixed-lever split-plot design, Resolution, Sub-plot factor, Whole-plot factor
PDF Full Text Request
Related items