Two High-level Factors Containing The Pure Effect Are Designed In The Mixed Horizontal Split Zone Of The Sub-area

Factorial design plays important role in the experiment. When the full experiment can not be carried out due to various reasons, we will choose the fractional factorial designs. If it is very difficult to change or control the levels of some factors, we will abandon the completely randomized experiment and choose fractional factorial split-plot (FFSP) designs to carry out the experiment.In many experiments, such as some physical and chemical experiments, the numbers of the levels of the factors are not equal to each other. Then we will choose mixed-level split-plot design. The high-level factors can be obtained from the low-level factors by the method of replacement.This thesis investigates the conditions of resolution III or IV mixed-level split-plot designs with two high-level factors in sub-plot such that the designs contain clear main effects or two-factor interaction components, and the corresponding construction methods of the designs. It is divided into five chapters.Chapter 1 contains four sections. The first section introduces the concept of facto-rial design and the full design. The second section introduces the construction method and the definition of two-level fraction factorial designs. The third section introduces the construction method and the definition of mixed-level fraction factorial designs. The forth section introduces the construction method and the definition of two-level fraction factorial split-plot designs.Chapter 2 introduces the properties and construction methods of fraction factorial split-plot designs in detail.Chapter 3 investigates the conditions of designs containing clear effects and the construction methods of the corresponding designs.Chapter 4 investigates the conditions of designs containing clear effects and the construction methods of the corresponding designs.Chapter 5 gives a brief summary of the thesis.