Planning Problems in Intermodal Freight Transport ...
Planning Problems in Intermodal Freight Transport ...
Planning Problems in Intermodal Freight Transport ...
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
able to <strong>in</strong>corporate non-monetary constra<strong>in</strong>ts such as schedule requirements and flatcar<br />
configuration restrictions <strong>in</strong> case different types of flatcars and trailers exist.<br />
A decision support system is constructed by Boardman et al. [63] to assist shippers <strong>in</strong><br />
select<strong>in</strong>g the least cost comb<strong>in</strong>ation of transportation modes (truck, rail, air, barge) between a<br />
given orig<strong>in</strong> and a correspond<strong>in</strong>g dest<strong>in</strong>ation. As an <strong>in</strong>dicator of cost average transportation<br />
rates for each transportation mode are used. This is a simplification of reality as there would<br />
normally be a cost difference between long haul truck and short haul drayage costs. Least-cost<br />
paths <strong>in</strong> the network are calculated by means of the K-shortest path double-sweep method.<br />
The software is <strong>in</strong>terfaced to a commercial geographic <strong>in</strong>formation system software package<br />
to assist the user <strong>in</strong> visualiz<strong>in</strong>g the region be<strong>in</strong>g analyzed.<br />
Ziliaskopoulos and Wardell [64] discuss a shortest path algorithm for <strong>in</strong>termodal<br />
transportation networks. The authors <strong>in</strong>troduce the concept of time dependency of optimal<br />
paths <strong>in</strong> their rout<strong>in</strong>g model. The time horizon is divided <strong>in</strong>to discrete <strong>in</strong>tervals. Also delays at<br />
switch<strong>in</strong>g po<strong>in</strong>ts, fixed time schedules of transport modes and movement delays or movement<br />
prohibitions are taken <strong>in</strong>to account. The algorithm computes optimal routes from all orig<strong>in</strong>s,<br />
departure times and modes to a dest<strong>in</strong>ation node and exit mode, account<strong>in</strong>g for the time-<br />
dependent nature of the arc travel times and switch<strong>in</strong>g delays, without explicitly expand<strong>in</strong>g<br />
the network. The computational complexity of the algorithm is <strong>in</strong>dependent of the number of<br />
modes. Computational time <strong>in</strong>creases almost l<strong>in</strong>early with the number of nodes <strong>in</strong> the network<br />
and the number of time <strong>in</strong>tervals.<br />
M<strong>in</strong> [65] focuses on the multi-objective nature of the modal choice decision. A<br />
chance-constra<strong>in</strong>ed goal programm<strong>in</strong>g (GP) model is constructed that best comb<strong>in</strong>es different