论文部分内容阅读
The transportation of crude oil from production fields to refineries where crude oil is processed,is a very important operation between the "upstream" and "downstream" fields of the oil industry.Today,because of the scarcity of crude oil and the rapid development of modem industry,oil demand has been increasing dramatically.However,most of the crude oil disperse in a few areas,such as Middle East.How to transport crude oil effectively from its origins to the demand points is a challenging issue in modern society.Major transportation modes of crude oil include pipeline and tanker transportation.Typical crude oil transportation cost is about 2.0 US$ per barrel,so any improvement to an existing crude oil transportation scheduling procedure will lead to a significant cost saving.Considering the storage capacity of each harbor and the transportation capacity on each route per period,an inventory routing model of the crude oil transportation problem is proposed in this paper,which is a mixed-integer nonlinear programming model.The problem is to determine the number of tankers that are rent or returned at harbors in each period and the number of tankers that are dispatched in each period and in each route at the minimum transportation cost (including pierage,fixed usage cost,operation cost,pipeline toll,canal toll,inventory cost,shortage penalty and rental cost),subject to the inventory balance constraints and capacity constraints.We consider the problem in a rolling horizon environment with multiple periods as well as the lead-time of transportation.Due to the complexity and the large size of the problem,the proposed mathematical model is too complicated to be solved exactly.A metaheudstic method GRASP is developed to fred near-optimal solutions of the model.Numerical test results of the method are provided.