需求不确定的海上溢油应急物资运送优化  

Optimizing emergency material delivery for marine oil spills in the face of uncertain demand

作　　者：陶宁蓉 詹慧萍 王世明[2,3] TAO Ningrong;ZHAN Huiping;WANG Shiming(College of Economics&Management,Shanghai Ocean University,Shanghai 201306,China;College of Engineering Science&Technology,Shanghai Ocean University,Shanghai 201306,China;Shanghai Engineering Research Center of Marine Renewable Energy,Shanghai 201306,China)

机构地区：[1]上海海洋大学经济管理学院,上海201306 [2]上海海洋大学工程学院,上海201306 [3]上海海洋可再生能源工程技术研究中心,上海201306

出　　处：《安全与环境学报》2025年第2期615-623,共9页Journal of Safety and Environment

基　　金：国家自然科学基金项目(41976194);上海市科学技术委员会资助项目(19DZ2254800);上海市“科技创新行动计划”软科学研究项目(23692102600)。

摘　　要：为了解决海上溢油事故发生时救援物资运送的问题,结合海上油膜的时变特性,考虑事故发生初期救援点位置变化和救援物资需求不确定,引入机会约束来确保满足救援点需求的可靠性,构建最小化环境污染和救援成本的双目标分布式鲁棒优化模型,并利用非支配排序遗传算法求解。结果表明,与随机规化和鲁棒优化相比,采用分布式鲁棒优化方法处理所研究问题得到的解总成本更低,分别减少了4.30%和3.95%。与不考虑需求扰动的模型相比,所建模型可使优化结果更稳定,环境污染、救援成本的变异系数可由10.77%、12.37%分别降至4.10%、5.83%。敏感性分析结果表明,在制定救援方案时,基于恰当的可靠性水平可有效降低救援成本和环境污染。Given that marine oil spills pose a significant threat to the marine environment and industry,it is essential to efficiently deliver emergency response materials in a manner that is environmentally sustainable.While previous studies have examined the time-varying characteristics of oil films influencing the placement of rescue points,they have not fully considered how these factors impact the demand for rescue materials.To address this issue,the study took into account the dynamic changes in the locations of rescue points and the uncertainties surrounding demand during the early stages of the accident.Initially,we simulated the time-varying characteristics of offshore oil films,including drift,diffusion,and evaporation,based on the meteorological and sea conditions at the oil spill site.This approach enabled us to predict the shifting locations of rescue points and assess the demand for rescue materials.Additionally,we took into account demand disturbances to enhance the resilience of the rescue plan.By examining the mean and variance of these disruptions,we implemented the opportunity constraint method to guarantee the effectiveness and reliability of meeting the material needs at rescue points.Next,we developed a Distributionally Robust Optimization(DRO)model aimed at minimizing both environmental pollution and rescue costs.By utilizing optimization techniques such as dual theory,we formulated the model into tractable counterparts.Additionally,we introduced a non-dominated sorting genetic algorithm to effectively address the bi-objective trade-offs within the model.In conclusion,a case study of the Penglai oil spill incident demonstrates the effectiveness of the model and offers valuable managerial insights.The findings indicate that:(1)In comparison to stochastic programming and traditional robust optimization approaches,the DRO method proves to be more adept at addressing the identified issue,resulting in a reduction of total costs by 4.30%and 3.95%,respectively.(2)Compared to the model that does not account

关 键 词：公共安全 海上溢油 物资调度 分布式鲁棒优化 需求不确定 救援点漂移 

分 类 号：X928.04[环境科学与工程—安全科学]