摘要
研究了多个受灾点突发灾情状况下应急物资的分配问题。首先基于对不同受灾点的供需相互转化的分析,建立了随机双层规划模型,其上层模型以极小化应急物资分配费用为目标,而下层模型则描述了不同受灾点的改善状况;然后采用贝叶斯更新来描述灾情演化过程中的突变状态;最后采用粒子群算法对该随机双层规划模型进行求解,算例求解结果表明在考虑灾情信息实时更新的情况下,所作出的动态决策使得整个救灾网络的灾情改善状况更好。
This article mainly focuses on the emergency material distribution under the condition of the un- expected situation. The article establishes the stochastic bi-level programming model by analyzing the transformation between supply and demand among different disaster sites to solve this circumstance. The upper level of the model aims at minimizing the rationality of the emergency supplies distribution to all af- fected areas, while the lower level describes the improvement of different disaster sites. The mutation situa- tion during the whole process of disaster is described by adopting the method of Bayesian updating. The ar- ticle further takes the particle swarm optimization to solve the stochastic bi-level programming model. A number of valid numerical examples demonstrate that the dynamic decisions may improve the situation of the entire disaster relief networks better when considering disaster information updated in real time.
出处
《安全与环境工程》
CAS
2015年第5期7-12,共6页
Safety and Environmental Engineering
基金
国家自然科学基金项目(71171129)
上海市科委科研计划项目(111510501900
12dz1124802)
上海市教委科研项目(11YZ137)
关键词
突发灾情
应急物资分配
双层规划模型
贝叶斯
粒子群算法(PSO)
unexpected disasters
emergency material distribution
bi-level programming model
Bayes
parti-cle swarm optimization（PSO）
