Models And Evolutionary Optimization For Emergency Logistics Scheduling In Disaster Relief

In recent years,there have been an increasing number of serious disasters including natural(earthquakes,floods)and man-made(terrorist attacks,chemical leakages),which have brought to the national major economic losses and casualties.The post-disaster emergency relief plays a vital role in reducing casualties and economic losses,and has become an important research topic.Emergency logistics is the process of planning,managing,and controlling the flow of those resources to provide relief to affected people in an event of disaster.Therefore,in disaster relief,it has great significance to build reasonable models and find the optimal emergency logistics scheduling plans.Aiming at the character of emergency logistics scheduling problems in disaster relief,this paper made the following studies:(1)A model for emergency logistics scheduling in disaster relief is built on the scenery of several suppliers with a variety of resources,several kinds of vehicles,and multiple disasters.In order to make the dispatch timely and effective,an objective is first designed to minimize the completion time of scheduling and the total unsatisfied time of the whole relief.Then,a multi-agent genetic algorithm(MAGA)is first proposed to solve this problem.To validate the effectiveness of the proposed algorithm,we make experiments on the case of the Chi-Chi earthquake in Taiwan,and the experimental results show that the convergence and the quantity of the solution obtained by our proposed algorithm are both better than that of genetic algorithm and memetic algorithm.(2)We design a new multi-objective model that considers both the total unsatisfied time and transportation cost.Then,a modified non-dominated sorting genetic algorithm II(m NSGA-II)is proposed to search for a variety of optimal feasible emergency scheduling plans for decision-makers.With the intrinsic properties of emergency logistics scheduling problems in mind,we design three repair operators to generate improved feasible solutions.Compared with the original NSGA-II,a local search operator is also designed for m NSGA-II,which significantly improves the performance.We conduct two experiments to validate the performance of the proposed algorithm,one is on the case of Chi-Chi earthquake in Taiwan,and the other is large-scale numerical experiment on synthetic data.(3)Considering the significant impact of uncertain parameters in real-life disaster relief,we design a scenario-based modeling method for uncertain road condition and a probing method for uncertain demand.Moreover,based on the multiobjective evolutionary algorithm based on decomposition(MOEA/D),we proposed a multi-objective optimization algorithm to search for a variety of optimal feasible emergency logistics scheduling plans for decision-makers.We conduct two experiments to validate the performance of proposed algorithm,one is on the case of Great Sichuan Earthquake in China,and another is on large scale numerical instances.