不确定信息下的灾后道路实时抢修鲁棒优化策略  

Robust optimization strategy for real-time road repair after disaster under uncertain information

作　　者：马倩乾 李树凯 杨立兴[1] 杨凯[1] Ma Qianqian;Li Shukai;Yang Lixing;Yang kai(School of Systems Science,Beijing Jiaotong University,Beijing 100044,China)

机构地区：[1]北京交通大学系统科学学院,北京100044

出　　处：《系统工程学报》2024年第3期360-372,共13页Journal of Systems Engineering

基　　金：国家自然科学基金资助项目(71942006,72288101).

摘　　要：针对突发事件后受损道路的抢修问题,建立了以抢修总成本最小为目标的多阶段道路抢修优化模型.进一步,考虑各个受损点抢修时间的不确定性,构建了基于情景集的灾后道路抢修鲁棒优化模型.为满足受损与抢修信息的预测不准确性和动态变化性,设计了一种基于多周期滚动优化方法的高效求解策略.最后通过数值实验发现,多阶段鲁棒优化抢修策略可以满足灾后道路抢修的不确定性和实时性要求.此外,决策者可根据实际需求选取合适的风险权衡因子,实现解鲁棒和模型鲁棒之间的权衡.In view of the problem of emergency repair of damaged roads,this paper develops a multi-period optimization model of emergency repair to minimize the total cost of emergency repair.Moreover,by consid-ering the uncertainty of emergency repair time,a robust optimization model of emergency repair is established by using scenarios.In order to meet the prediction inaccuracy and dynamic variation of damage and emergency repair information,an efficient solution strategy is proposed based on multi-period rolling-horizon optimization method.Finally,from numerical experiments,we canfind that the multi-period robust optimization emergency repair strategy can satisfy the uncertainty and real-time requirements of road repair after disaster.In addition,decision makers can select appropriate risk trade-off factors according to actual needs to realize the trade-off between solution robustness and model robustness.

关 键 词：道路抢修 鲁棒优化 滚动优化 不确定性 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]