Optimization of emergency evacuation network based on hybrid decision-making method
Natural disasters cause huge economic losses,affect the lives of many people,and severely damage the environment.The suddenness and destructiveness of earthquakes and other disasters result in great and diverse demand for refuge.However,materials are often not delivered quickly and effectively due to the inefficient matching of supply and demand of emergency resources.Rapid response to urgent needs immediately after earthquakes through an efficient emergency evacuation network is vital to alleviate the impact of disasters in affected areas.Designing an emergency evacuation network for the preparedness and response stages improves relief efforts and service quality and supports disaster management operations during large-scale earthquakes.However,so far,this issue has received scant attention.This paper comprehensively considers the location of different types of emergency shelters,the allocation of evacuees,and the distribution of supplies under layered evacuation demands.Based on the post-disaster three-layer evacuation structure framework,an emergency evacuation network with a recursive mechanism is proposed,including:1)the number of evacuation victims and the material demand forecast;2)shelter location and victim evacuation;3)material distribution.A hybrid decision support method,including fuzzy clustering and multi-objective dynamic programming models,is proposed to optimize the number of shelters selected,evacuation distances,and material fulfillment rates.Using Chengdu City,Sichuan Province as an example to deal with emergency evacuation and rescue plans for large-scale earthquake disasters,the process of evacuation and allocation of evacuees is analyzed.The index fitting method and the idea of safety stock are used to predict emergency evacuation demand.Combined with the influence of the material distribution center on the location of emergency shelters,a multi-objective optimization model is constructed to formulate plans for the evacuation of victims from multiple demand points to different types of emergency shelters and a high-quality non-dominated solution is obtained with the help of the weighted ε-constraint method.According to the real-time material supply and demand situation,the urgency of material demand for different types of emergency shelters is determined by the fuzzy clustering method,and the multi-period dynamic distribution model of emergency materials is constructed to complete material distribution activities fairly and efficiently.The experimental results show that the proposed method can fairly and quickly solve the problems of site selection,evacuation of victims,and material distribution in the early post-disaster period and ensure the orderly and efficient operation of the emergency evacuation.The applicability of the method proposed in this paper in different scenarios is verified by evaluating different metrics such as the average and longest evacuation distance,the average and minimum material satisfaction rate,and the average and longest material transportation distance.Considering the design of the emergency evacuation network in the first three days after the earthquake,including the site selection of emergency shelters,material distribution,and evacuation of victims,the designed network nodes mainly include demand points,emergency shelters,and material distribution centers,integrating resource supply and demand sides.First of all,the main needs of the victims are basic life services or psychological and medical assistance services.They are divided into two categories.Considering factors such as the number of residents in demand points,the proportion of victims allocated to shelters,the proportion of victims who need different services,the number of different types of evacuees,and the changes in the number of materials required in the early post-earthquake period are analyzed,and a prediction model for the number of evacuees is constructed.The safety stock concept is used to forecast the demand for emergency supplies.Secondly,combined with the characteristics of demand-driven rescue in the emergency evacuation network and the idea of multi-layer coverage,a model of emergency evacuation site selection and evacuation of victims based on hierarchical evacuation demand is proposed,considering multiple optimization objectives:first,the total evacuation distance of victims after the earthquake is the minimum.The second is to minimize the number of selected emergency shelters,and systematically allocate evacuees with different needs to emergency shelters with different functions for the characteristics of weak economic levels in disaster emergency response.To find the Pareto front of the related problem,the weighted e-constraint method converts the bi-objective function into a single objective function to generate a non-dominated Pareto set.It is convenient to adjust the weight of the objective function to adapt to the weak economy of the emergency evacuation demand in the early stage after the earthquake and in the middle and later stages,when the disaster situation is stable and under the condition of giving priority to the convenience of evacuation,moderately pursuing the economic goal of reducing disaster losses.Next,to solve the problem of shortage of emergency supplies and vehicles in the first three days after the earthquake,the priority of material distribution is divided to transport the materials quickly and accurately to the victims most in need of assistance.The sources of emergency supplies include the territory and the rear.Determining the priority attribute vectors is related to each emergency shelter from the aspects of the materials obtained by the shelter,the structure of the victims,and the post-disaster environment.Combined with the real-time material supply and demand situation,the fuzzy clustering method divides the priority of material distribution for refugees and disaster victims.In different periods,combined with the time-varying weight of emergency shelters,the total satisfaction rate of time-varying emergency supplies demand is maximized.Finally,the data of a district in Chengdu,Sichuan Province,are selected to evaluate the proposed method,which verifies the feasibility and effectiveness of the model,and three indicators to quantitatively assess the applicability of the proposed method in different scenarios are designed,especially for rapid response to disaster victims and in terms of quickly responding to the different evacuation demands of the victims and coordinating multiple sources of supplies.Numerical calculation results show that the joint optimization of disaster evacuation and material distribution process,consideration of the fairness of the evacuation process,setting material distribution priority and a minimum satisfaction rate,and ensuring material transportation capacity and material reserves can all be used as decision support tools for emergency evacuation networks,especially when quickly responding to emergency rescue needs in large-scale disaster areas.Finally,it is hoped the hybrid method proposed in this paper can solve actual emergency evacuation and rescue problems.In the future,further research will be conducted on the uncertain problems and dynamic optimization involved in improving the operational performance of emergency evacuation networks.
万方数据
维普
