Recently, the research community has focused on the problem of efficient movement of freight by rail since there has been a tremendous increase in worldwide container trade. For example, there is an estimation that the Ports of Los Angeles and Long Beach (San Pedro Bay Ports), which are one of the busiest ports in North America, will double their current cargo by year 2020. This raises the demand for freight trains and leads to congestion in rail operation.; Several train studies in the past focused on solving the train scheduling problem by simplifying and omitting many realistic characteristics such as trackage configurations or speed restrictions. However, we incorporate these characteristics into our models so we can obtain a solution that can be easily applied to actual railway operations more efficiently. These characteristics include multiple trackage configurations, multiple train types, multiple train lengths, multiple speed limit restrictions, and incorporation of acceleration and deceleration rates.; The purpose of this research is to optimize the train travel time for general rail networks. We formulate a nonlinear integer mathematical model with an objective function to minimize the travel time. Since the model is computationally intractable to solve optimally, we present a heuristic based on a genetic algorithm to solve the problem. We benchmark our algorithm against standard procedures for train scheduling. We study both the static and dynamic scheduling problems since each of those problems has their own unique characteristics.
展开▼
