本文目的在于利用混合血样DNA 信息，估计混合血样贡献者人数。本文提出了两种提高现有估计方法准确率的办法：第一，选择最优DNA 位点组合；第二，向原有的极大似然估计模型中，加入gamma 分布拟合的等位基因峰高信息。结果显示，估计混合血样的人数时，选用的DNA 位点越多，估计越准确；加入峰高信息来估计人数，的确会提高原有极大似然估计算法的准确率，但是选用几种不同的分布来拟合峰高信息，准确率相差不大。

This essay aims to estimate the number of contributors to DNA forensic mixtures. We put up with two ideas to improve the efficiency of current literature. The first is to choose the best combination of DNA locus to estimate the number of contributors, because according to our survey, some of the DNA locations are almost unhelpful in estimating the number of forensic DNA mixtures. The second idea is to add allele peak height probability into the Maximum Likelihood Estimation Model, and we choose gamma distribution to fit the distribution of allele peak height. It is showed that the more DNA locations we use the more efficient the algorithm is. And adding peak height has greatly improved the original algorithm, however, it is not important whether we choose gamma distribution as our fitting model or not.