计算机科学
弹性(材料科学)
人口
时间范围
应急管理
运筹学
分类
遗传算法
调度(生产过程)
备份
数学优化
交通拥挤
流量网络
瓶颈
最优化问题
路径(计算)
功能(生物学)
供求关系
多目标优化
杠杆(统计)
意外事件
战略规划
桥接(联网)
进化算法
作者
Xinrong Li,Hui Hu,Lu Chen,Jiao Zhao,YaQing Zeng
标识
DOI:10.1177/03611981261427739
摘要
Emergency supply distribution networks face significant resilience challenges during large-scale disasters because of hub congestion and demand–supply mismatches. This study addresses this issue by proposing and comparing two congestion management strategies for hybrid hub-and-spoke rail–road intermodal networks: a waiting versus path redistribution strategy using backup hub mechanisms. A multi-objective optimization model was constructed to maximize network resilience, minimize transportation time, and reduce costs. Resilience was measured by the demand gap weighted by demand urgency. A rolling horizon optimization framework is established to address the temporal dynamics of disaster relief operations. A Q-learning-enhanced Non-dominated Sorting Genetic Algorithm II (NSGA-II) algorithm is developed to solve the optimization problem, constructing a 16-dimensional state space by integrating objective function values and population diversity metrics for intelligent local search. Using severely affected areas from the Wenchuan earthquake as a case study, the experimental results demonstrate that the improved algorithm reduces average objective function values by 17.75%, 49.82%, and 19.92%, respectively, compared with the standard NSGA-II. Incorporating demand urgency factors reduces the material shortage index by 53.41%, better reflecting humanitarian priorities. By comparing the average function values across Periods 1–6, the first four periods are suitable for the path reallocation strategy, while the subsequent two periods should adopt the “continue waiting” strategy. The study provides actionable insights for emergency managers in optimal strategy selection in disaster relief operations.
科研通智能强力驱动
Strongly Powered by AbleSci AI