Optimization Of Drop And Pull Transport Scheduling Considering The Trailer Storage Capacity Of Customs

As an advanced transportation organization pattern,Drop and Pull Transport(DPT)has been widely used in western countries with the advantages of low cost,high operation efficiency and being environment-friendly.Due to the restrictions of domestic transportation policy,technical standards and the practical scheduling pattern etc,DPT has not yet been widely applied in China.Among all the restrictions,lacking of scientific and reasonable scheduling method is a major bottleneck restraining the development of DPT.Therefore,studying DPT scheduling in-depth and using the theory of scientific method to guide practice and service practice become a hot research topic needed to be solved.In the operation of DPT,tractors and trailers can be flexibly separated and coupled.Through the reasonable scheduling between them,waiting time of the tractors can be greatly reduced,so as to improve the productive efficiency of the tractors.A high complexity is implied in tractor-and-trailer scheduling because the two transport resources need to be scheduled collaboratively.To take full advantage of DPT’s potential collaborative-scheduling flexibility,this paper proposed a new transport pattern of full load DPT based on the collaboration of tractors and trailers.Supporting the proposed transport pattern,a transport network composing of a depot and a certain number of customers is built as a scheduling framework in which one trailer can be matched with different tractors.On the basis of dividing the job behavior of the tractor and the trailer,we formulated a linear mathematical model with the residual time matrix as core to describe the scheduling behaviors of tractors and trailers.And the scheduling scheme is disintegrated to the task dispatching problem for tractors and the coordination problem between tractors and trailers.A hybrid GA(genetic algorithm)and greedy-based heuristics are employed to solve the bi-level problem.Large-scale experimental results demonstrated that the model and algorithms proposed in this paper get an efficient and practical applications in DPT problem.