Study On Mixed Fleet Scheduling Problem With Time Windows Under Different Deliveries

As the main pillars of logistics, transport act as the connection of merchants and customers, with the functions of improving economic efficiency of logistics, optimizing logistics, improving service and reducing costs. Thus, it gradually becomes the breakthrough of the headspring of the profit. In recent years, the proportion of transport cost in the logistics cost increases gradually with the fuel price rising and logistics market competition intensifying. Therefore, the effective measure and key to compress logistics cost and improve the logistics service level is how to improve the enterprise’s own management level, carry out effective transport organization and dispatch scientifically. Traditional fleet scheduling problem require that vehicles in the team are homogeneous, but there are few teams that all vehicles in it are homogeneous in practice. At the same time, there are customers’time windows and high idle rate.Therefore, it has important practical significance studying on mixed fleet scheduling problem with time windows under different deliveries.First, operational mechanism of fleet scheduling problem is described from overview, constituent elements and classification. Against this basis, operational mechanism and algorithm of mixed fleet scheduling problem are studied. Second, Mix Vehicle Routing Problem(MVRP) is solved by Tabu Search. The model is built on the basis of detailed description of MVRP, the original scheduling scheme is produced through Saving Algorithm and improved through Tabu Search. Then, Mix Vehicle Routing Problem with Time Windows is solved by Threshold Accepting. According to the characteristics of MVRPTW and MVRP model, original scheduling scheme is produced through improved saving algorithm, and route in it is eliminated. And the original scheduling scheme is improved by searched locally through Threshold Accepting. Finally, Mix Vehicle Routing Problem with Split Deliveries is solved through Scatter Search. MVRPTWSD model is established according to the character of the problem, and original scheduling scheme is produced through improved saving algorithm, and problem is solved through Scatter Search. Based on the detailed description of mixed fleet scheduling problem, this problem is studied by successive progressive from the view whether customer time window is existed or not and whether distribute in patch or not. It has important significance to improve the logistics economic benefits for enterprise and implement logistics modernization.