本文主要以非对称分布中泊松分布为研究对象, 探讨了其参数的稳健估计方法. 在本文中, 作者以截断似然估计为基础, 结合Cizek的工作, 提出了适用于泊松分布参数的一种稳健估计方法. 该方法避免了事先选取截断比例的麻烦, 通过数据自身的信息给出在平均似然最大准则下的最优截断比例. 在文中的模拟部分, 分别就未受污染和受污染的泊松分布数据进行了模拟, 得到了不错的效果.

In this paper, the author takes the Poisson distribution as representative of unsymmetrical distributions and studies the robust estimation method of the parameter. The author suggests a robust estimation method for the parameter in Poisson distribution based on the work of Cizek and the method of maximum trimmed likelihood estimation. The method mentioned in this paper avoids to choose the trimmed proportion of data in advance, and gives the optimum proportion by the information of data under the criterion of maximum average likelihood function. In simulation part, the data of Poisson distribution which be contaminated and not be contaminated has carried out a simulation respectively, has got good effect.