复杂灾害情景下应急资源配置的随机规划模型

A stochastic programming model for relief resources allocation problem based on complex disaster scenarios

提出基于区域灾害系统理论来构建复杂灾害情景,用于描述突发事件的复杂性和高度不确定性.基于复杂灾害情景建立了一个两阶段随机规划模型,进行应急设施的定位决策、应急物资的库存决策和不同灾害情景下应急物资分配预案的制定.使用新的编码方式,将第一阶段的选址-库存模型转化为一个无约束非线性优化模型,使用自适应免疫克隆选择文化算法进行求解.对于第二阶段的应急物资分配模型,则提出了"势能抵消算法".最后基于四川地震带的相关数据给出了算例,验证了模型及算法的有效性与可行性.

Complex disaster scenarios are constructed based on regional disaster system theory to describe complexity and high degree of uncertainty of emergencies. Then a two-stage stochastic programming model is established based on disaster scenarios to solve emergency facilities location problem, relief supplies in- ventory problem and relief supplies allocation planning problem in different disaster scenarios. To use adaptive immune clonal selection cultural algorithm, the location-inventory model in the first stage is con- verted into an unconstrained nonlinear optimization model by new encoding. And a potential cancellation algorithm is proposed to solve the relief supplies allocation model in the second stage. Finally a numerical example is given based on the Sichuan earthquake zone to verify the effectiveness and feasibility of the model and algorithm.