The Research on the Problem of Emergency Logistics Route Optimization
|School||Three Gorges University|
|Course||Applied Computer Technology|
|Keywords||Ant Colony Algorithm emergency dispatch TSP Vehicle Route Problem section|
In the paper the research aimed to critical problems is done, one is multi-depot singledisaster multi-material emergency dispatch problem, another is single depot multi-disasteremergency logistics route optimization problem.On the multi-depot single disaster multi-material emergency dispatch problem, annew multi-depot and multi-material emergency dispatching algorithm consideringcontinuous consumption is proposed, in which both the ability of a depot in satisfying thedemand of the disaster area and the effect of choosing a depot upon choosing the next onein the process of choosing depot one by one are considered, and a series ofnon-inferiority candidate depots are obtained, then they are forced to try to participate inthe rescue to search for the least depots solution.Before the research on single depot multi-disaster emergency logistics routeoptimization problem is done, its sub-problems are researched respectively, and twoalgorithms are proposed. One is the improved Ant Colony Algorithm on solving TSPproblem. The route found by Saving Algorithm is made to be its initial best route, whichmakes the improved ant algorithm can be optimized in a high starting point; for graspingsome local characteristics of the best route, more complete transcendental knowledge forprobability choosing formula is provided; through strengthening the relative guiding effectof the best route found in the last iterations to accelerate the convergence of the improvedant colony algorithm; through the application of the local best route tabu strategy to avoidthe algorithm to fall into the local best route. The algorithm is compared with the Max-MinAnt Colony Algorithm, the results show that the improved algorithm can converge muchfaster, and the quality of the solutions is much higher.Another is the two stage algorithm on solving the Vehicle Route OptimizationProblem, firstly several groups are obtained, and every group is comprised by severalmaterial demanded points. The demand of every group can be transported by only onevehicle. Secondly the improved Ant Colony Algorithm is applied in the optimization ofevery vehicle’s transportation route, then a better transportation route is obtained bycombining the sub-routes together.At last, a emergency logistics route optimization algorithm which considers thedemands of disasters and the time spent in transportation between two disasters by vehiclesas sections. When the Algorithm proceeds the demands of disasters and the transportation time of vehicles, the combined distribution functions of the demands of disasters which areserved by a vehicle and the transportation time of the vehicle are obtained. Then it iscomputed that whether the vehicle can satisfy the demand and the time limit of nextdisaster the vehicle will go to by a high possibility, If the answer is “yes”, then the vehicleis forced to choose the disaster through possibility choosing. Otherwise, if any disastercan’t be satisfied by the vehicle by a high possibility, the vehicle will return to the depot,then let another vehicle to complete the mission of material transportation of the leftdisasters. As the improved Ant Colony Algorithm is introduced in the algorithm, a solutionwill be obtained in which the total cost is much lower after the algorithm is over.Eventually, the effectiveness of the algorithm is verified through an example.