内集-外集模型的计算机仿真检验

基于信息分配方法建立的内集 -外集模型 ,用于计算可能性 -概率风险 ,可以表达概率估计的模糊性。实践证明 ,用这一模型计算出来的自然灾害模糊风险能对减灾方案进行合理的筛选 ;实验证明 ,我们还可以用这一模型替代专家依据给定的样本进行模糊概率估计。本文用计算机仿真技术 ,在期望值的意义上 ,检验这一模型的可靠性。仿真结果显示 ,用内集 -外集模型计算出来的可能性 -概率分布的均值甚至比传统直方图分布的均值更靠近真实概率分布的期望值。这一研究表明 ,内集-外集模型用于自然灾害模糊风险评估 ,在期望值的意义上是一个可靠的模型。

Computer simulation tests of interior-outer-set mode

The interior outer set model, based on the method of information distribution, can represent the imprecision of probability estimation when it is used to calculate a possibility probability risk. A practice proved that reasonal disaster reducing scheme can be sifted according to the fuzzy risk of the natural disaster calculated by this model. An experiment proved that instead of experts this model can be used to estimate fuzzy probabilities based on given samples. This paper uses a computer simulation technique to test reliability of the interior outer set model in the meaning of the expected value. Simulation results show that the expected value of a possibility probability distribution calculated by using this model is more closer to the real expected value than that from a histogram estimation. This study demonstrates that, in the meaning of the expected value, the interior outer set model is reliable for calculating the fuzzy risk of the natural disaster.