Study On Multi-objective Vehicle Routing Problem In Urban Emergency Logistics

China is a large country, all kinds of emergencies and disasters occur frequently in recent years, which have gravely infringed on the citizens' lives, given an enormous impact to people's economic life. Dealing with public emergency is a complicated systematic project, the supplies of the relief materials timely is the key to increase the efficiency of rescue work. Because of this, it is necessary to establish a scientific and efficient emergency logistics system, and then we can supply the relief material in the shortest time to minimize the serious consequences caused by the public emergency. Emergency Logistics Vehicle Routing Problem is the key to the optimization problems of emergency logistics system. It must consider minimizing not only the transportation time of relief material but also the total system cost, so it is a multi-objective optimization problem.Firstly, this paper reviews the emergency logistics and the vehicle routing problem in emergency logistics at home and abroad. then, investigates the difference between logistics and emergency logistics, the vehicle routing problem and the vehicle routing problem in emergency logistics, the vehicle routing problems with sided time windows, traffic time in emergency logistics, advantages and disadvantages of commonly used algorithms for vehicle routing problems. on this basis, two multi-objective optimization models are proposed either taking into account minimizing the total time and total system cost or taking into account maximizing the satisfaction with the system time and total system cost. As the decision-making between the Emergency Logistics VRP and the general logistics VRP is evidently different, the emergency logistics VRP lays more emphasis on the efficiency of time, so this article uses a weight coefficient transformation method to give greater weight to time-related objective function. This paper uses an approximate sub-function with setting a peak for the travel time, and updates the speed in real-time to solve the reach time to the hazards area.Traditional solution for multi-objective optimization problem is not sensitive to the Pareto boundary, if you want to obtain Pareto optimal solution, you must run the optimization process many times,and each optimization process is relatively independent, the results are often inconsistent. Genetic Algorithms is potentially effective in dealing with the intermittent and multi-peak nature of the objective function. This article designs an efficient multi-objective genetic algorithm to solve the model, and discusses the encoding, selection mechanism, crossover and mutation processes of it for the Emergency Logistics VRP in detail. Finally, we verify the validity of the algorithm with simulation, which provides the decision support tools for emergency management personnel in emergency transport of relief material after the outbreak and assists them to develop contingency logistics projects.