Perfect homogeneous rail traffic: A quick efficient genetic algorithm for high frequency train timetabling

This paper treats a particular train timetabling problem (TTP) in order to enable a high-frequency schedule of the future intercity network of London. The treated rail traffic is homogeneous because all trains have similar characteristics, as it is generally the case of the subway and intercity rail traffic. Overtaking opportunities between trains are studied; in order to evaluate whether overtaking contributes to improve the train's frequency. This study allows us to introduce a new class of train traffic that is called “perfect homogeneous”. Perfect homogeneous rail traffic can be considered as a one way train track. We define High Frequency Train Timetabling Problem (HF-TTP) without overtaking. The problem is formulated, using an analogy with the traveling salesman problem. A genetic algorithm to quickly define the sequence of trains for reducing the period is presented.