计算机工程与应用 ›› 2025, Vol. 61 ›› Issue (16): 348-356.DOI: 10.3778/j.issn.1002-8331.2406-0089

• 工程与应用 • 上一篇    下一篇

基于鲁棒优化的伤员分类救治中心选址-运输问题研究

王尧,丁祥海,俞武扬   

  1. 1.杭州电子科技大学 管理学院,杭州 310018
    2.湖州学院 经济管理学院,浙江 湖州 313000
  • 出版日期:2025-08-15 发布日期:2025-08-15

Research on Location and Transportation Problem of Casualty Classification and Treatment Center Based on Robust Optimization

WANG Yao, DING Xianghai, YU Wuyang   

  1. 1.School of Management, Hangzhou Dianzi University, Hangzhou 310018, China
    2.School of Economics and Management, Huzhou College, Huzhou, Zhejiang 313000, China
  • Online:2025-08-15 Published:2025-08-15

摘要: 应急医疗服务是震后救援活动的重要组成部分,但伤员人数的不确定性增加了决策的困难性。为了减少灾后损失,基于鲁棒优化考虑伤员数量的不确定性,同时结合伤员分类及救治中心分型的思想,构建一个以伤员满意度最大化和救援工具运营成本最小化为目标的双目标鲁棒优化模型,并采用NSGA-Ⅱ(non-dominated sorting genetic algorithm-Ⅱ)设计求解方法。结果表明:扰动比例和不确定水平的组合对救治中心的选址和伤员的分配方案有显著影响,不同程度的伤员人数波动对目标函数值会有不同的影响,决策者可根据自身对本次地震灾害的判断,得到不同的选址分配方案,并与实际情况相结合做出科学合理的决策。

关键词: 选址运输, 分类救治, 双目标规划, NSGA-Ⅱ, 鲁棒优化

Abstract: Emergency medical services are an important part of post-earthquake rescue activities, but the uncertainty of the number of casualties increases the difficulty in decision-making. In order to reduce post-disaster losses, this paper considers the uncertainty of the number of casualties based on robust optimization, combines the ideas of the casualty classification and treatment center classification, constructs a dual-objective robust optimization model to maximize the satisfaction of casualties and minimize the operation cost of the rescue tools, and uses the NSGA-Ⅱ (non-dominated sorting genetic algorithm-Ⅱ) to design the solution method. The results show that the combination of disturbance ratio and uncertain level has a significant impact on the site selection and the distribution of casualties. The variation of the number of casualties to different degrees will have a different effect on the target function value. Decision-makers can get different location distribution plans according to their judgment of the earthquake disaster, and make scientific and reasonable decisions combined with actual conditions.

Key words: location transportation, classified treatment, dual-objective programming, NSGA-Ⅱ, robust optimization