Abstract: Recently, China has suffered frequent floods. In order to transport emergency materials after disasters, a three-layer network system of emergency logistics is usually adopted. After the implementation of the distribution plan, materials are uniformly deployed to the transfer stations through the distribution center and then distributed to the disaster areas. During the transportation process, people compare the arrival time and amount of distributed materials in their own area with those in other affected areas to understand the psychological pain and sense of imbalance suffered by victims due to insufficient materials. Therefore, a reasonable planning of material deployment and storage center location is crucial to improving the psychological loss of victims. In this study, the cost function of psychological loss, the absolute cost function of psychological suffering and the relative cost function of psychological suffering are established to quantify the psychological loss of victims. At the same time, a two-level planning model for the location of cross-regional emergency materials deployment is proposed, which takes into account factors such as various materials types, transportation modes, regional disaster level, material urgency level and satisfaction rate, and aims to minimize the psychological cost of victims, select temporary transit stations, and formulate material deployment plans. In the decision-making process, the senior management considers the psychological loss and the absolute psychological torment cost of the affected people, chooses the temporary reserve station as the material transfer station, and arranges the material from the distribution center to the nearest emergency material transfer station, and adopts an appropriate transportation mode. The junior management considers the relative cost of psychological torment of the affected people, and arranges the distribution and transportation of materials. To solve this model, a discrete problem-solving method based on multiverse optimization algorithm is proposed. By introducing nonlinear convergence factor and spiral motion strategy, the problem of fast convergence in the early stage of the algorithm is improved, and the diversity and accuracy of the solutions are ensured. Finally, taking Wuhan city as an example, we compare the experimental results of genetic algorithm, multiverse optimization algorithm and improved multiverse optimization algorithm, and verify the effectiveness of the proposed algorithm. The study also points out that the results such as material satisfaction and time of delivery are significantly correlated with the degree of destruction and vulnerability of the disaster area, taking into account the psychology of the disaster victims. In addition, the satisfaction rate of emergency materials with low urgency demand is significantly lower than that of emergency materials with high urgency demand. This study has enriched the research content of emergency logistics and provided quantitative decision support for decision makers to optimize the deployment of emergency materials and the location of transfer stations when considering the psychological loss of disaster victims. In the future research, we should strengthen the recycling and utilization of available resources and waste materials in the emergency logistics network, and establish a two-way distribution logistics network model of emergency materials after disaster considering the psychological loss of victims.

Key words: emergency material location-scheduling, bi-level programming, improved multi-verse optimization algorithm, psychological losses of the victims

摘要： 突发洪涝灾害发生后灾民的心理处于易损状态,应急物资调配方案制定以及配送中心选址对灾民的有效救援有重大影响。本文在软时间窗、物资供应量小于需求量约束下,同时考虑多物资种类、运输方式、区域受灾程度以及不同物资急迫性、最低满足率等因素,建立跨区域应急救援物资调配——选址双层规划模型。根据模型特点,提出一种改进多元宇宙优化算法的求解离散问题算法。引入非线性收敛因子和围绕螺旋移动策略,保证解的多样性和精确度。最后以武汉市作为实例研究对象,对比遗传算法、基本多元宇宙算法与改进多元宇宙算法实验结果,对模型与算法的可行性与有效性进行验证。研究结果表明,该方法可以为决策者提供科学决策支持。

关键词: 应急物资调配——选址, 双层规划, 改进多元宇宙优化算法, 灾民心理损失