The logistics routes allocation discusses optimal routing from origin to destination through distribution center （DC） on condition that both transport and distribution cost should be taken into account. The problem ...The logistics routes allocation discusses optimal routing from origin to destination through distribution center （DC） on condition that both transport and distribution cost should be taken into account. The problem can be solved by the iterative non-linear programming （INLP）, in which the transport cost and distribution cost are assumed to be linear and non-linear, respectively. The method works well in most situations. However, when the distribution cost predominates in the total cost, the method falls, and the solution given by the method is not a global minimum but a local minimum. Further study reveals that the INLP method is still a kind of transport routing method like vehicle routing problem （VRP）, and the failure of the method must happen when the distribution cost is a major one. On such a condition, further computation on other extreme points, which physically means forcing all routes to pass through one DC one by one, should be carried out. By comparing values on these extreme points, the global optimal solution can be got. The method has both theoretical and practical meaning. In theoretical field, it might force us to seek new method; in practice, it reminds us to do such kind of check when the transport distance is short and warehousing work is major that often happens in local consolidation center or de-vanning center.展开更多
This paper addresses the scheduling problem involving batch processing machines, which is Mso known as parallel batching in the literature. The presented mixed integer programming formulation first provides an elegant...This paper addresses the scheduling problem involving batch processing machines, which is Mso known as parallel batching in the literature. The presented mixed integer programming formulation first provides an elegant model for the problem under study. Fhrthermore, it enables solutions to the problem instances beyond the capability of exact methods developed so far. In order to alleviate computational burden, the authors propose MIP-based heuristic approaches which balance solution quality and computing time.展开更多
基金the National Science Foundation,Ministry of Education and Science, Japan (No. 17330089)
文摘The logistics routes allocation discusses optimal routing from origin to destination through distribution center （DC） on condition that both transport and distribution cost should be taken into account. The problem can be solved by the iterative non-linear programming （INLP）, in which the transport cost and distribution cost are assumed to be linear and non-linear, respectively. The method works well in most situations. However, when the distribution cost predominates in the total cost, the method falls, and the solution given by the method is not a global minimum but a local minimum. Further study reveals that the INLP method is still a kind of transport routing method like vehicle routing problem （VRP）, and the failure of the method must happen when the distribution cost is a major one. On such a condition, further computation on other extreme points, which physically means forcing all routes to pass through one DC one by one, should be carried out. By comparing values on these extreme points, the global optimal solution can be got. The method has both theoretical and practical meaning. In theoretical field, it might force us to seek new method; in practice, it reminds us to do such kind of check when the transport distance is short and warehousing work is major that often happens in local consolidation center or de-vanning center.
文摘This paper addresses the scheduling problem involving batch processing machines, which is Mso known as parallel batching in the literature. The presented mixed integer programming formulation first provides an elegant model for the problem under study. Fhrthermore, it enables solutions to the problem instances beyond the capability of exact methods developed so far. In order to alleviate computational burden, the authors propose MIP-based heuristic approaches which balance solution quality and computing time.
