Minimizing congestion on railway networks

The railway transportation is becoming more and more important to the whole transportation system. The railway transportation is cost-effective, energy-efficient and safe. The railway is a major method of transportation of freight and passengers in many countries. Given the fact that the railway industry is facing rapid increasing demand, better planning and scheduling tools are needed to effectively manage the scarce resources. In this thesis, we focus on developing train scheduling and train dispatching algorithms that minimize the congestions on railway networks.;The first part of this thesis studies the freight train scheduling problem. We develop optimization-based approaches for scheduling of freight trains. Two mathematical formulations of the scheduling problem are first introduced. Two heuristics based on mixtures of the two formulations are proposed. The proposed algorithms are able to perform better than two existing heuristics such as a simple look-ahead greedy heuristic and global neighborhood search algorithms, in terms of total train delay. For relatively large networks, two algorithms based on the idea of decomposition are proposed. We use two sample networks to test our proposed decomposition algorithms. The two sample networks are 33.9-mile and 49.3-mile long, respectively. Our decomposition algorithms are shown to outperform the two existing algorithms.;The second part of the thesis studies the train dispatching problem for double-track and triple-track railway segment. The regular dedicated dispatching policy dedicates one track for trains travelling in one direction. If trains are travelling at different speeds, the faster train will catch up with the slower train thus be delayed. We first look at the problem where there are only two train speeds in the network, we propose a switchable dispatching policy for the double-track segment to reduce the chance that a faster train being blocked by a slower train. We use queueing theories to derive the delay functions. The new switchable dispatching policy is shown to significantly reduce the train delays compared to a dedicated policy. We also study the effect of having a crossover at the middle of the track segment. The switchable policy is then extended to the triple-track segment. Several variations of the switchable policy for triple-track segment are studied. Finally, we study the double-track segment dispatching policy for multiple train speeds.