Construction Of Split-plot Designs With General Minimum Lower Order Confounding

In practice,when the level of some factors in the experiment is di cult to change or control,the fractional factorial split-plot（FFSP）design is a good choice for experimenters.It has been widely used in many fields such as industry and agriculture because of its advantages of saving experiment cost.For the fractional factorial（FF）design,Zhang et al.^[26]introduced a new aliased effect number pattern（AENP）for evaluating FF designs based on the effect hierar-chy principle.Based on AENP,the general minimum lower order confounding（GM-C）criterion is proposed to select the optimal FF design.When the experimenter has a prior knowledge of the importance ranking of each factor in the experiment,the GMC design is superior to other optimal designs.The GMC criterion is one of the recognized effective criterions used to choose FF design,it is also widely used in the selection of FFSP design.In this thesis,2^{（n₁+n₂）-（m₁+m₂）}GMC-FFSP designs in the range 9N/32+1≤n₁+n₂≤5N/16 are investigated.The construction methods of 2^{（n₁+n₂）-（m₁+m₂）}GMC-FFSP designs for the cases where n₁=1 with m₁=0,n₁=2 with m₁=0,n₁=3 with m₁=0,n₁=4 with m₁=1,and m₂=n₂-1 are worked out.