中文摘要：

混合偏差是在已有的偏差下提出的一种新的偏差,它克服了中心化L2偏差和可卷L2偏差的一些不足.混合偏差作为部分因子设计的均匀性测度,寻找它的精确下界非常重要.获得了二水平设计混合偏差的一个新的下界,数值例子说明它比已有的下界在某些设计中更加精确.

英文摘要：

The mixture discrepancy is a new discrepancy based on existing discrepancy,which overcomes some weakness of the centered and wrap-round L2-discrepancies.As the measure of uniformity of fractional factorial designs,it is very important to look for the accurate lower bounds to the mixture discrepancy.This paper gives a new lower bound to the mixture discrepancy in two levels fractional factorial designs.The new lower bound is better than existing lower bounds in certain factorials designs.Finally,some examples are given to illustrate the results.