Name: PATRICIA COSTA LOUZADA LOPES
Type: MSc dissertation
Publication date: 12/12/2017
Advisor:
Name |
Role |
|---|---|
| RODRIGO DE ALVARENGA ROSA | Advisor * |
Examining board:
| Name |
Role |
|---|---|
| JOÃO LUIZ CALMON NOGUEIRA DA GAMA | Internal Examiner * |
| LEANDRO COLOMBI RESENDO | External Examiner * |
| RODRIGO DE ALVARENGA ROSA | Advisor * |
Summary: Meeting demand for rail cargo transport is related to the wagons and locomotives distribution process for making up trains in order to transport cargo, which directly influences railroad's operational costs. Planning the distribution of empty wagons is important because most wagons on Brazilian railroads returns empty from the unload rail yard to the cargo rail yard. In case of locomotives they eventually travel alone to meet demands in other rail yards. Both situations increase the railroad operational costs because a train made up only of empty wagons, or a locomotive traveling alone, adds engine driver and diesel costs, and does not bring any revenue for the railroad. Thus, this dissertation proposes a Integer Programming mathematical model for planning the integrated distribution of wagons and locomotives, WHERE they are transported using the slack in the capacity of traction of loaded trains that are already scheduled to circulate in the railroad to meet empty wagons demands of and locomotives requests in the yards aiming to minimize the total distribution costs. The proposed model is distinguished by integrating wagons and locomotives distribution, which has not been found in the literature yet. In order to guarantee the balance between offer and demand, it is proposed the introduction of gap variables to represent the demands that have not been met by the actual quantity offered or by the limitation in the traction capacity available in circulating trains. In the case of using these variables to obtain the optimal solution, it means that the demand in that yard and time will not be met, indicating that the resource distributor must take an action. Test instances based on real data from the Vitória a Minas Railroad (EFVM) were solved to optimality using the solver CPLEX 12.6 and the results were quite consistent with the operational use to support decision making in planning and distribution of wagons and locomotives over rail yards to meet the demand for making up trains. In EFVM real case instances, the proposed model used fewer trains made up exclusively of empty wagons to distribute wagons and locomotives than those actually created by the Railroad Operational Control Center in the period analyzed.
Keywords: Wagons and Locomotives Distribution Planning. Railroad Operation. Railroad Transport. Integer Programming.
