基于行为运筹的灾后伤员救援车辆及手术调度优化研究

In recent years, the disasters occurred frequently and caused heavy casualties in the world. It is urgently needed to carry out meaningful research on the optimal distribution of disaster victims and the scheduling problem of surgical operations. The disaster injuries and surgeries, different from normal ones, have specific features such as complicate injuries, high time sensitivity, large volume, etc., which urgently need an efficient solution to the injury distribution and surgery scheduling requirements. Aiming at these new problems, this proposal carries out research work by using the combination of theoretical analysis and empirical research, data calculation and simulation experiments, quantitative research and qualitative analysis based on behavioral science, scheduling and optimization theory. Firstly, this topic collects data via questionnaire, lab experiments and field survey, etc., and researches injury characteristics of casualties and injury identification methods. Then, it is proposed to further build the vehicle distribution and operation scheduling models by integrating different learning effect models and scene features and to study the coordination of post-disaster wounded rescue vehicles and surgical scheduling problems. Finally, the project will design intelligence algorithm system and carry out the empirical study and simulation research. Especially, the proposed models are to be built from a new perspective by considering the rescue attendants' behavior endowment effect. The outputs of this research are expected to enrich the theory system and improve the operations management in rescuing casualties, thus providing the theoretical basis and methodological guidance for post-disaster casualty rescue.

近年来，全球灾害事件频繁发生，造成了大量人员伤亡，开展针对灾后伤员分送和手术调度问题的研究刻不容缓。灾后伤员伤情及手术具有不同于普通伤员的新特点，灾后伤员伤情复杂、时效性强、批量较大，对伤员分送和手术调度的需求更为迫切。针对这些新问题，本课题基于行为科学、排序论和最优化理论，运用理论分析和实证研究相结合、数据计算和仿真实验相结合、定量研究和定性分析相结合的方法，通过文献检索、现场调研、问卷调查等手段获取数据，研究灾后伤员伤情特征和伤情识别分类方法，构建针对不同干扰要素、不同学习效应模式和具体场景特征的灾后伤员运送车辆和手术调度模型，研究灾后伤员救援车辆及手术调度优化问题，设计群智能优化算法体系，开展实证研究和仿真研究。本项目考虑救护人员学习行为，从新的视角提出面向重大灾害事件的伤员分送和手术调度模型，预期成果有助于完善灾后伤员救援管理的理论体系，为灾后伤员救护提供理论基础和方法指导。

近年来，地震、台风、洪涝、疫情等突发灾害事件频发，给人民的生命财产造成了严重损失，也对经济社会发展产生了重大影响。灾后伤员伤情及手术具有不同于普通伤员的新特点，灾后伤员伤情复杂、时效性强、批量较大，对伤员分送和手术调度的需求更为迫切。. 针对这些新问题，本项目基于行为运筹学、不确定理论、模糊理论、干扰管理理论、组合优化理论、智能优化理论，运用理论分析和实证研究相结合、数据计算和仿真实验相结合、定量研究和定性分析相结合的方法，综合考虑伤员伤情、需求变化、道路交通情况等多种影响因素，分别构建了伤员分级分类模型、伤员救护车辆调度模型、伤员手术调度模型，并根据模型的特性，设计了基于萤火虫算法、灰狼优化算法、鲸鱼优化算法等群智能优化算法的高性能求解算法。. 本项目构建的数学优化模型，考虑了伤员的分级救治、伤情的恶化、道路损坏、伤员到达时间以及手术持续时间的不确定性等现实约束，模型与实际场景吻合度高，鲁棒性强，求解的决策方案提高了救援效率且将灾害造成的人员伤亡程度降低到最小。设计的算法性能优越，在保证求解质量的前提下，大大地提升了计算效率，适用于求解类似大规模组合优化问题。. 本课题研究成果不但丰富和发展了重大灾害情景下应急管理和伤员救助的理论、方法与技术体系，还能够为各级应急管理决策部门在有限的时间、空间和资源约束条件下，系统化开展快速、高效、有序的应急救援工作，实施科学应急、高效应急、精准应急，提供重要的理论基础和决策参考。

内容获取失败，请点击重试