2fi最优行列设计的构造
Construction of 2fi-optimal row-column designs
AI总结:
本文针对高素数水平填补了2fi最优行列设计的构造空白,提出了任意奇素数水平全因子及任意素数水平部分因子2fi最优设计的理论构造方法。
摘要:
能够提供所有主效应的无混杂估计以及最大数量两因子交互效应(2fi's)的行列因子设计被称为2fi最优。由于其在工业或物理实验中的广泛应用,该问题近年来备受关注。此前已有文献提出了2fi最优的二水平和三水平全因子及部分因子行列设计的构造方法。然而,针对高素数水平的结果尚未实现。本文中,我们推进了这些构造工作,给出了针对任意奇素数水平s和任意参数组合的s^n全因子2fi最优行列设计的理论构造,以及针对任意素数水平s和任意参数组合的s^{n-1}部分因子2fi最优行列设计的理论构造。
英文摘要:
Row-column factorial designs that provide unconfounded estimation of all main effects and the maximum number of two-factor interactions (2fi's) are called 2fi-optimal. This issue has been paid great attention recently for its wide application in industrial or physical experiments. The constructions of 2fi-optimal two-level and three-level full factorial and fractional factorial row-column designs have been proposed. However, the results for high prime level have not been achieved yet. In this paper, we develop these constructions by giving a theoretical construction of $s^n$ full factorial 2fi-optimal row-column designs for any odd prime level $s$ and any parameter combination, and theoretical constructions of $s^{n-1}$ fractional factorial 2fi-optimal row-column designs for any prime level $s$ and any parameter combination.
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。