用计算机仿真技术检验自然灾害模糊风险模型

自然灾害模糊风险是用模糊集对自然灾害风险的一种近似表达．本文建议了一种计算机仿真方法，用来检验计算模糊风险的模型是否可靠．假定真实的概率风险可以用某个概率密度函数来表达，当用某种模型依据小样本来估计这个函数时，必然会存在误差．设给定概率密度函数的其望值为Ｅ， 用传统模型估计得到的概率密度函数的期望值为Ｅ，用某种模糊数学方法计算出来的模糊概率分布的期望值为Ｅ．如果Ｅ与Ｅ的差小于Ｅ与Ｅ的差，则说明模糊风险模型比较可靠．本文给出了计算Ｅ的公式，对计算机仿真实验进行设计．并给出了产生随机数的３个重要程序．

A computer simulation method to test fuzzy risk model of natural disaster

The fuzzy risk of natural disaster is an approximate representation showing risk with fuzzy sets . In this paper, a computer simulation method is suggested to test the reliability of a model used to calculate a fuzzy risk. It is supposed that a real probability risk can be represented by using a probability density function. When a model is used with a sample to estimate the function, there must exist error. If the expected value of the real function is E, the expected value of the estimated function calculated by using a traditional model is E, the expected value related to a fuzzy mathematics model is E and the error between E and E is smaller than that between E and E, then the fuzzy model is more reliable than the traditional model. In this paper, the formula to calculate E is given, the computer simulation experiment is designed, and three programs to produce random numbers are provided as well.