贝叶斯决策法是最常见的以期望为标准的分析方法.传统贝叶斯方法的基本原理是,在不完全情报下,对部分未知的状态用主观概率估计,利用贝叶斯公式对发生概率进行修正,最后根据修正概率和期望值做出最优决策.贝叶斯方法直接应用于城市地震应急决策时存在两个问题:将灾害事件作为单一事件考虑及直接计算不同地震级别下各方案的效用.为提高决策的科学性,文中对贝叶斯方法进行改进,将修正过的后验概率与次生灾害处于不同级别的概率相结合,得到更精确的概率;并将各方案的效用分解为各方案对次生灾害的子效用,进行综合后得到总效用,由此计算出的总效用更精确、具体.将改进后的方法应用于地震应急决策,通过算例分析得出最优方案,说明了该方法的有效性和实用性.%The Bayesian decision-making is commonly used analysis approach taking the expectation as the standard. The basic principle of the traditional Bayesian method is to estimate the probability of the occurrence of the unknown state with the subjective probability and to use the Bayesian formula to modify the probability of occurrence and make the optimal decision according to the modified probability and expected value. When Bayesian method is applied directly to the urban earthquake emergency decision making,there are some prob-lems such as treating disaster event as a single event and calculating the utility of each solution directly under different earthquake levels. Therefore,it is needed to improve Bayesian method. Firstly,the modified posterior probability and probability of secondary disasters in different levels is combined to obtain more accurate probability. Then the utility of each solution can be decomposed into the sub utility of the secondary disasters. Consequently,the improved method is applied to the earthquake emergency decision making. It demonstrates that the modified method is effective and practical.
展开▼
