Research On Optimization Method For Vehicle Scheduling Of Logistics Enterprise

With the rapid development of intelligent transportation and e-commerce in recentyears, modern logistics presents a trend characterized by informatization andsocialization. Growing business volume not only brings unlimited opportunities for thevigorous development of logistics enterprises, but raises new requirements for logisticstransport capacity and efficiency. The critical issue logistics enterprises face is to realizethe profit maximization through the improvement of logistics transport efficiency andits transportation at the lowest cost. Focusing on Vehicle Routing Problem (VRP) oflogistics enterprises, this thesis starts with the description and analysis of Ant ColonyOptimization (ACO) algorithm to solve relevant problems, and then put forward fourimproved strategies for its defects. Besides, the superiority of the improved ACOalgorithm is proved through simulation analysis. On this basis, the thesis designs anddevelops a Vehicle Scheduling System.The thesis firstly elaborates the definition, elements and classification of VehicleScheduling Problem, in which emphasizes Capacitated VRP and Vehicle RoutingProblem with Time Windows (VRPTW). Based on this, the practical significance andmathematical description of various constraints are analyzed and explained in detail,thus two distinctive mathematical models are established respectively aiming at thesetwo problems.Secondly, the thesis focuses on the study of ACO to solve VRP which has thepositive feedback mechanism and high stability. In view of the deficiencies oftraditional ant colony like slow search speed as well as easy to fall into local optimalsolution, four improvements are proposed involving the construction of initial solutionbased on Sweep algorithm, parallel search of certainty and exploratory, dynamicpheromone updating and key parameter optimization. By solving the problem related totravelling salesman, the improved ACO has proved its high efficiency and stability forthe shortest route. Then, the improved ACO is further applied to solve Capacitated VRP and VRPTW.The implementing steps of solving the two kinds of problems are designed respectivelyand simulation analysis are done by adopting different measure samples. Comparingwith traditional ant colony algorithm, it has obvious superiority whether in thecalculation of optimal solution, search speed or robustness.Finally, the thesis designs architecture and function of Vehicle Scheduling Systemof logistics enterprises, realizing the integrated management to information refers tocustomer, parking lot, goods, order for goods as well as vehicles; furthermore,automatically generation of vehicle scheduling tasks with time windows and mapdisplay of vehicles’ shortest route are both achieved. Through the calculation andanalysis of actual order for goods, distribution the system generates is rational andeffective, which is of great guiding significance to vehicle scheduling of logisticsenterprises.