Capacity management of railroad lines based on queuing systems and higher-dimensional Markov chains
Efficiency Is Key
RWTH researchers are helping to make the most efficient use of the German rail network.Copyright: © Felix Schreiber
Those who do not like highway congestion or air traffic control strikes often decide to travel by train. However, as many people know from bitter experience, rail transportation follows its very own laws. Unfortunately, train cancellations and delays tend to be the order of the day, especially during peak travel season. A sensible use of the railroad network is, however, of decisive importance not only for passenger but also for freight traffic. Professors Anke Schmeink and Nils Nießen are conducting interdisciplinary research to find a solution to this problem, thanks to an ERS Seed Fund Grant.
Nils Nießen, a passionate train passenger himself, is head of the Institute of Transportation Sciences at RWTH Aachen University. In his research, he investigates at a wide range of issues, from the construction of station facilities or railroads to the necessary safety technology or the control of transportation operations.
Electrical engineering professor Anke Schmeink, on the other hand, approaches the project from a mathematical perspective. In her research she optimizes networks, for example future mobile communication networks, which will probably increasingly determine our everyday life in the years to come. This includes, for example, intelligent risk predictions or communication between machines in Industry 4.0.
Together with several doctoral candidates, the two professors have developed a computer model to optimize waiting times in driving operations. This is particularly important for junctions, branch-off points, and single-track line sections, as rail traffic has to flow through a tight bottleneck here. Since, for example, a single-track line can only be travelled along in one direction while the train at the other side has to wait, delays can occur more frequently, which has negative affects the entire rail network. In order to minimize or even completely avoid such backlogs, it is particularly important to coordinate the times each train is scheduled to run or wait. The ideal distribution can be calculated using special algorithms. Temporary events, such as line closures, can be included in the calculations or timetables optimized.
The resulting computer model should make it easier to find the ideal operation strategy for stations or sub-networks and to plan new rail facilities. According to the two professors, it is even possible to avoid building expensive, superfluous facilities in the first place. This is particularly important in the case of so-called "overthrow structures", i.e. bridges and underpasses, which have a particularly high financial impact. Different infrastructure solutions for a given route occupancy can be compared by simulating the resulting waiting times. In this way it can be determined whether it would be sufficient to optimize rail operations on the modelled line section or whether new construction is actually necessary. In the end, a suitable software must be developed for users. However, it will probably take a few more years for the current model to be ready.
"Our findings can certainly be transferred to other transportation systems such as trams, subways, aircraft, or even cars," says Nießen. The LUKS software tool was developed within the company even before the ERS project, with DB Netz AG already using it for some time now, he proudly adds.
Follow-up projects to the ERS Seed Fund are currently being planned and a British university has already expressed keen interest in sharing knowledge on the topic. Considering that the use of rail transportation is constantly increasing and switching to rail travel also has the huge potential of reducing CO2 levels, this field of research is clearly very important.
Put simply, this means that investments in infrastructure are also an indispensable investment in our future in times of increasing mobility.