|Title||Automatic assignment algorithms for loading double stack rail cars|
|Publication Type||Journal Article|
|Year of Publication||1995|
|Authors||Jahren CT, Rolle SS, Spurgeon LE, Palmer RN, Newman RR, Howland DL|
|Journal||Transportation Research Record|
|Keywords||Algorithms, Containers, Heuristic methods, Loads, Monte Carlo method, Performance evaluations, Railroad cars|
The development of two automatic suggestion algorithms (ASAs) for loading containers onto double-stack railcars is described. A container-oriented ASA (COASA) considers each arriving container and selects a loading position (LP) on the train. A location-oriented ASA (LOASA) considers each loading position and selects a container from the arrival pool (the containers at the terminal entrance queue). Both approaches use heuristics to improve train loading quality. A well-loaded train has a high load factor, low center of gravity, and uniform load distribution along the length of the train. Metrics were developed for each of these measures of performance. The loading strategies are tested using the Monte Carlo method based on historical container arrival data and typical train configurations. The performance of the LOASA improves when the pool size is increased, with the greatest improvement occurring when the pool size increases from one to two. For pool sizes greater than two, the COASA and the LOASA have similar performance. A simplified algorithm also was tested and evaluated. That algorithm produced load factors similar to the LOASA and the COASA, but did not perform as well according to the other metrics.