Research on Transportation Resources Scheduling Based on Network Drop-and-pull Transport
Abstract:: Drop-and-pull transportation can shorten logistics chain, improve transportation efficiency, and promote intensification of the transportation industry. However, currently the drop-and-pull transporta⁃ tion practice is still at a low development level in China and in the application of the drop-and-pull transporta⁃ tion mode, traditional vehicle route optimization methods are lacking in aspects of realistic matching and con⁃ straint expression, etc. In this paper, in view of the vehicle dispatching problem in the drop-and-pull transpor⁃ tation mode, we replaced the more common VRP with the service network design, and established a mixed in⁃ teger programming model to meet the transportation conditions while minimizing the dispatching cost, there⁃ by achieving the goal of cost reduction and efficiency improvement. In view of the intercity trunk network scenario, given the same premises, we established the drop-and-pull transportation model aimed at minimizing the sum of the fixed tractor and trailer cost, the tractor operat⁃ ing cost and the cargo cost, and a traditional transportation model with the minimum sum of the fixed truck cost and the operating cost of the trucks and the cargoes as the objective function, and proceeded to compare the two models. Then taking a third-party logistics enterprise as the research object, we had a numerical study respectively using a sample of 20 orders (small scale sample) and a sample of 50 orders (medium scale sample) and calculated the two parameters of average time window length and task distribution concentration in the model solution process: (1) When solving the small scale sample problem, both models could obtain optimal so⁃ lution within 1 minute. When solving the medium scale sample problem, the drop-and-pull transportation model with its all linear constraints performs better in terms of solution time than the traditional transportation model some of whose constraints are nonlinear. When the average length of transportation tasks increases, the cost of network drop-and-pull transportation does not change much, but the traditional transportation pro⁃ cess requires more vehicles to complete the tasks, causing the cost to increase significantly; when the distribu⁃ tion of the tasks deviates from the mid-time of the transportation cycle, the costs under both modes will in⁃ crease significantly, but at this time, the number of vehicles required for the drop-and-pull transportation mode will increase remarkably, resulting in a reduction in the cost optimization effect. (2) In the small scale sample problem, we further analyzed the performance of the scheduling plans, laid out three transportation states of full load, empty load and waiting for service, with their time percentage representing the proportion of the different states in the scheduling plan, and found that the advantage of the network drop-and-pull trans⁃ portation practice is reflected in higher full-load percentage, which can achieve dual optimization purpose of reducing vehicle quantity and lowering routing costs. The numerical example thus verified the feasibility and effectiveness of the drop-and-pull transportation model. Based on the results obtained, in practical applica⁃ tions, enterprise decision-makers can decide whether to upgrade to network drop-and-pull based on the pa⁃ rameters of the transportation network and the characteristics of the transportation tasks, and can rely on the speed advantage of the transportation practice in small and medium-scale problems to find the optimal schedul⁃ ing solutions.
Engineering