外文翻译--多路程车辆排程问题之门槛接受启发式演算法(编辑修改稿)

外文翻译--多路程车辆排程问题之门槛接受启发式演算法(编辑修改稿)内容摘要：

VRP, the Period VRP (PVRP), and VRP with Backhauls, etc. Some academics study on the problems which bine more than two kinds of extended properties. The length of Taiwan from north to south is only about 394 kilometer and most of the plains in Taiwan lie in the west. Therefore, most industries are located at the west and the distance of transportation is shorter than the other country. The basic VRP supposes that each available vehicle can only be used once in the planning period such as an eight hours daily workinghour. It means that each vehicle can only be assigned one route to distribute. But the total distribution time of one planning route in Taiwan may be less than a half of the planning period. Hence, we can use fewer vehicles to finish the distribution plan if each vehicle can be assign more than one route. It is more conform to the environment of Taiwan and the real world distribution applications. This extend problem is called the Vehicle Routing Problem with Multiple Use of Vehicles (VRPM) or the Multitrip Vehicle Routing Problem (MTVRP). But the VRPM in many real world applications is more plex than the VRP. The VRP is only one level of the plex routing problem in the real world applications. For example, the VRPM that allows for more than one route assigned to each vehicle is a three levels problem of decision making. At the top level is an assignment problem about assigning customers (nodes) to vehicles, and the second level problem is the VRP in each vehicle. At last, the TSP in each single route of the VRP is the third level problem. Because of its multiple levels of decision problems relating to the VRP, this kind of problem is called a Multilevel Vehicle Routing Problem (MLVRP). However, the research of the VRPM is not much in the literature. Homes et al. (1989) did an empirical study of the real distribution problem related to the VRPM. In 1990, Fleischmann suggested using bin packing to solve the VRPM in his working paper. Then, Taillard et al. (1996) proposed a heuristic based on the Tabu Search method (TS) and bin packing for solving the VRPM. Brandao and Mercer (1998) presented a heuristic based on the TS for solving the VRPM. There is some research related to the VRPM and bined with other properties such as time。