Laura Galli

Solution of the Train Platforming Problem

In this paper we study a general formulation of the train platforming problem, which contains as special cases all the versions previously considered in the literature as well as a case study from Rete Ferroviaria Italiana, the Italian Infrastructure Manager. In particular, motivated by our case study, we consider a general quadratic objective function, and propose a new way to linearize it by using a small number of new variables along with a set of constraints that can be separated efficiently by solving an appropriate linear program. The resulting integer linear programming formulation has a continuous relaxation that leads to strong bounds on the optimal value. For the instances in our case study, the solution approach based on this relaxation produces solutions that are much better than those produced by a simple heuristic method currently in use, and that often turn out to be (nearly) optimal.

Railway Rolling Stock Planning: Robustness Against Large Disruptions

In this paper we describe a two-stage optimization model for determining robust rolling stock circulations for passenger trains. Here robustness means that the rolling stock circulations can better deal with large disruptions of the railway system. The two-stage optimization model is formulated as a large mixed-integer linear programming (MILP) model. We first use Benders decomposition to determine optimal solutions for the LP-relaxation of this model. Then we use the cuts that were generated by the Benders decomposition for computing heuristic robust solutions for the two-stage optimization model. We call our method Benders heuristic. We evaluate our approach on the real-life rolling stock-planning problem of Netherlands Railways, the main operator of passenger trains in the Netherlands. The computational results show that, thanks to Benders decomposition, the LP-relaxation of the two-stage optimization problem can be solved in a short time for a representative number of disruption scenarios. In addition, they demonstrate that the robust rolling stock circulation computed heuristically has total costs that are close to the LP lower bounds. Finally, we discuss the practical effectiveness of the robust rolling stock circulation: When a large number of disruption scenarios were applied to these robust circulations and to the nonrobust optimal circulations, the former appeared to be much more easily recoverable than the latter.

A tutorial on non-periodic train timetabling and platforming problems

In this tutorial, we give an overview of two fundamental problems arising in the optimization of a railway system: the train timetabling problem (TTP), in its non-periodic version, and the train platforming problem (TPP). We consider for both problems the planning stage, i.e. we face them from a tactical point of view. These problems correspond to two main phases that are usually optimized in close sequence by the railway infrastructure manager. First, in the TTP phase, a schedule of the trains in a railway network is determined. A schedule consists of the arrival and departure times of each train at each (visited) station. Second, in the TPP phase, one needs to determine a stopping platform and a routing for each train inside each (visited) station, according to the schedule found in the TTP phase. Due to the complexity of the two problems, an integrated approach is generally hopeless for real-world instances. Hence, the two phases are considered separately and optimized in sequence. Although there exist several versions for both problems, depending on the infrastructure manager and train operators requirements, we do not aim at presenting all of them, but rather at introducing the reader to the topic using small examples. We present models and solution approaches for the two problems in a didactic way and always refer the reader to the corresponding papers for technical details.

Gap inequalities for non-convex mixed-integer quadratic programs

Laurent and Poljak introduced a very general class of valid linear inequalities, called gap inequalities, for the max-cut problem. We show that an analogous class of inequalities can be defined for general non-convex mixed-integer quadratic programs. These inequalities dominate some inequalities arising from a natural semidefinite relaxation.

Delay-Constrained Shortest Paths: Approximation Algorithms and Second-Order Cone Models

Routing real-time traffic with maximum packet delay in contemporary telecommunication networks requires not only choosing a path but also reserving transmission capacity along its arcs, as the delay is a nonlinear function of both components. The problem is known to be solvable in polynomial time under quite restrictive assumptions, i.e., equal rate allocations (all arcs are reserved the same capacity) and identical reservation costs, whereas the general problem is

Delay-Robust Event Scheduling

\mathcal {NP}

Optimal Joint Path Computation and Rate Allocation for Real-time Traffic

NP-hard. We first extend the approaches to the equal rate allocation (ERA) version to a pseudo-polynomial Dynamic Programming one for integer arc costs and a FPTAS for the case of general arc costs. We then show that the general problem can be formulated as a mixed-integer Second-Order Cone (SOCP) program and therefore solved with off-the-shelf technology. We compare two formulations: one based on standard big-M constraints and one where Perspective Reformulation techniques are used to tighten the continuous relaxation. Extensive computational experiments on both real-world networks and randomly generated realistic ones show that the ERA approach is fast and provides an effective heuristic for the general problem whenever it manages to find a solution at all, but it fails for a significant fraction of the instances that the SOCP models can solve. We therefore propose a three-pronged approach that combines the fast running time of the ERA algorithm and the effectiveness of the SOCP models, and show that it is capable of solving realistic-sized instances with high accuracy at different levels of network load in a time compatible with real-time usage in an operating environment.

Flexible dynamic coordinated scheduling in virtual-RAN deployments

Train Routing is a problem that arises in the early phase of the passenger railway planning process, usually several months before operating the trains. The main goal is to assign each train a stopping platform and the corresponding arrival/departure paths through a railway station. It is also called Train Platforming when referring to the platform assignment task. Railway stations often represent bottlenecks and train delays can easily disrupt the routing schedule. Thereby railway stations are responsible for a large part of the delay propagation in the whole network. In this research we present different models to compute robust routing schedules and we study their power in an online context together with different re-scheduling strategies. We also design a simulation framework and use it to evaluate and compare the effectiveness of the proposed robust models and re-scheduling algorithms using real-world data from Rete Ferroviaria Italiana, the main Italian Railway Infrastructure Manager.