Sponsor:

Innovative Scheduling

Research Team:

Joseph Geunes, Guvenc Sahin, Siddhartha Mohapatra

Universities Involved:

University of Florida

Start Date:

08/31/04

End Date:

08/31/05

Summary:

The objective of this project is to model and analyze the current yard operations, develop new optimization-based policies for humping, classification, and assembly, and develop simulation-based decision support system to compare different policies
Railroads want to improve their efficiency of yard operations. An efficient yard operations planning will lead to improvements in the overall railroad system. Thus, yards planning decisions are one of the important aspects of rail operations. Yard operations consist of a series of processes dependent on each other. At a yard, the classification and assembly operations involve the separation of cars from the inbound trains, the sorting of these cars according to their blocking plans and combining them to form the outbound trains. In this series of processes, since the cars are transferred from one inbound train to another outbound train should pass through classification and assembly processes, the challenge lies in dealing with the coordination among the classification and assembly operations.

A hump yard consists of a set of receiving tracks where the inbound trains arrive, a set of classification tracks where the cars from inbound trains are humped and then combined to form the outbound blocks, and a set of departure tracks where the outbound trains wait for their assembly to be completed. All three types of tracks are capacitated in terms of the number of cars lying on them. The main problem encompasses the following interrelated problems:

(1) How the inbound trains are allocated onto receiving tracks and in which sequence they are humped?
(2) How to arrange classification track occupation?
(3) In which sequence the outbound blocks should be built and when they are assembled to the outbound trains?

Each of the above problems includes finite capacity resource allocation coupled with ordering decisions. Thus, they are highly combinatorial in their nature. Moreover, the dependency among these problems makes the overall problem even more difficult to solve.

With the tools for planning yard operations, railroads aim to:

(1) increase the utilization of receiving tracks and departure tracks,
(2) maximize the utilization of classification tracks,
(3) minimize the average dwell time per car,
(4) minimize the switching done to assemble the outbound trains, and
(5) minimize the number of missed outbound connections.

When these objectives are realized, in addition to better utilization of the above mentioned resources, it will also result in reduced yard jobs. Thus, overall results will help the railroad increase the capabilities of the yards.