定量因子设计的字次型

发布时间：2018-05-24 21:39

本文选题：字长型 + 字次型　；参考：《苏州大学》2015年硕士论文

【摘要】：确定比较不同试验设计好坏的标准对于选择合适的设计来说是至关重要的.当所考虑的试验仅包含定性因子或二水平定量因子时,最小低阶混杂设计是值得推荐的.但对于多水平定量因子试验来说,仅仅采用最小低阶混杂准则来选择设计是不够的.为比较定量因子设计的好坏,本文提出了字次型的概念,并给出了最小低次混杂的判别准则.首先,对于正则设计,考虑其定义关系子群及其对偶码的码字分布特征,引出正则设计字次型和最小低次混杂的的定义.通过对组合同构的正则设计L9(33)和L27(313)字次型的具体计算,说明了字次型在对具有相同参数的设计的优劣比较中的实际应用.而对于非正则设计,我们利用编码理论中的MacWilliams恒等式,给出了字次型的推广定义及计算公式.同样,通过对两个L1s(37)设计字次型的计算,说明了字次型作为衡量非正则设计好坏准则的可行性.
[Abstract]:Determining the criteria for comparing the quality of different test designs is essential for choosing the right design. The minimum low-order hybrid design is recommended when only qualitative factor or two-level quantitative factor is included in the experiment. But for multi-level quantitative factor test, it is not enough to select the design by using the minimum low order hybrid criterion. In order to compare the quality of quantitative factor design, this paper puts forward the concept of word type and gives the criterion of minimum low order mixing. Firstly, for regular design, the definition of relation subgroup and codeword distribution of dual codes are considered, and the definitions of regular design word subtype and minimum low order hybrid are derived. The practical application of the word order type in the comparison of the advantages and disadvantages of the design with the same parameters is explained by the concrete calculation of the regular design L9N33) and the L27H313) word type of combinatorial isomorphism. For irregular designs, we use the MacWilliams identity in coding theory to give the generalized definition and calculation formula of word type. In the same way, through the calculation of two L _ 1s ~ (-37) design word types, the feasibility of word order type as a criterion for evaluating irregular design is illustrated.
【学位授予单位】：苏州大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：O212.6

【相似文献】