Modern control systems based on Moving Block allow for shorter headways and higher capacity on existing railway infrastructure. At the same time, few algorithms for optimal routing on networks equipped with such modern control systems exist. Previous methods rely on Mixed Integer Linear Programming (MILP) and face a trade-off between model size and accuracy, especially considering comparably complex and nonlinear headway constraints as well as train dynamics. With this work, we propose a complementary approach based on A*. Under a reasonable and easy assumption on train driver behavior, we propose a solution encoding and state space that is flexible concerning the choice of search algorithm and the modeling detail. The applicability is showcased on a small benchmark set. The implementation is available open-source as part of the Munich Train Control Toolkit (MTCT) on GitHub at https://github.com/cda-tum/mtct. 2012 ACM Subject Classification Applied computing -> Transportation
We propose an algorithm for computing alternatives to the shortest path in a road network, based on the speed-up technique CCH (customizable contraction hierarchy). Computing alternative paths is a well-studied problem, motivated by the fact that route-planning applications benefit from presenting different high-quality options the user can choose from. Another crucial feature of modern routing applications is the inclusion of live traffic, which requires speed-up techniques that allow efficient metric updates. Besides CCH, the other speed-up technique supporting metric updates is CRP (customizable route planning). Of the two, CCH is the more modern solution with the advantages of providing faster queries and being substantially simpler to implement efficiently. However, so far, CCH has been lacking a way of computing alternative paths. While for CRP, the commonly used plateau method for computing alternatives can be applied, this is not so straightforward for CCH. With this paper, we make CCH a viable option for alternative paths, by proposing a new separator-based approach to computing alternative paths that works hand-in-hand with the CCH data structure. With our experiments, we demonstrate that CCH can indeed be used to compute alternative paths efficiently. With this, we provide an alternative to CRP that is simpler and has lower query times. 2012 ACM Subject Classification Theory of computation -> Shortest paths; Mathematics of computing -> Graph algorithms; Applied computing -> Transportation
Setting the ticket prices is a crucial decision in public transport. Its basis, relevant for all related questions, such as dynamic prices or prices for different passenger groups, is the underlying fare strategy. Popular fare strategies are based on zones or on distances. Transitions from one fare strategy to another occur frequently, e.g., if public transport operators are joined to a larger association, or if structural decisions in a region have taken place. In this paper we report practically relevant issues when a fare structure should be changed to a distance tariff, a problem frequently arising when a ticket system based on mobile devices is introduced. We present mixed-integer linear programs for finding the parameters of a distance tariff, analyze rounding properties, and reflect how the change in revenue for the operator and the number of highly affected passengers can be controlled. Additionally, we evaluate the developed models experimentally. 2012 ACM Subject Classification Applied computing -> Transportation; Mathematics of computing -> Mixed discrete-continuous optimization
The Target Visitation Problem (TVP) combines the Traveling Salesman Problem and the Linear Ordering Problem, and thus serves as a natural model for route planning applications where both the travel costs and the order of the sites to visit matter. More precisely, in addition to the costs that apply for the selected links connecting two subsequently visited sites, the relative urgency of visiting one site before another is quantified and taken into account. In this article, we present refined integer linear programming formulations for the TVP, along with clarifications and extensions regarding the description of the polytopes associated with their feasible solution sets by a minimal set of linear equations and facet-defining inequalities. The practical effectiveness of exploiting the proposed improvements by means of a branch-and-cut algorithm is demonstrated in a computational study. In addition, we report the optimal values for some previously unsolved instances. 2012 ACM Subject Classification Mathematics of computing -> Combinatorial optimization; Applied computing -> Transportation; Mathematics of computing -> Linear programming; Mathematics of computing -> Integer programming
While a shift from individual transport to public transport reduces greenhouse gas emissions, public transport itself also consumes a non-negligible amount of energy. Acceleration processes have a high part in that, especially in urban transportation networks where stops are not far from each other. Express lines which skip stops hence use less energy than a vehicle on a normal line on the same route. Additionally, they increase the attractiveness of public transport by reducing travel times. In this paper, we introduce the express line planning problem ELP which extends the well-known line planning problem by the additional planning of express lines and which stops they skip. The problem is stated in a bicriteria setting minimizing the passengers travel time and the energy consumption of the public transport system. We investigate the problem's complexity and develop two different MIP formulations and show their equivalence. The models are tested numerically on medium sized instances. 2012 ACM Subject Classification Applied computing -> Transportation; Applied computing -> Decision analysis
Ridepooling becomes more and more popular and providing comfortable and easy-to-use transportation (nearly as taxi rides) is known to motivate passengers to use public transport. In this paper we develop a model for strategic planning of ridepooling. Here we decide in which regions ridepooling should be offered and what capacities are needed, neglecting the operational details of dial-a-ride planning. We use this model for integrating ridepooling and line planning, and analyze the integrated model theoretically and numerically. Our experiments show the potential of the approach. 2012 ACM Subject Classification Applied computing -> Transportation; Applied computing -> Decision analysis
We introduce a first-of-its-kind efficient, exact algorithm for the dynamic taxi-sharing problem with single-transfer journeys, i.e., a dispatcher that assigns traveler requests to a fleet of shared taxi-like vehicles allowing transfers between vehicles. We extend an existing no-transfer solution by collecting all viable pickup and dropoff vehicles for a request and computing the optimal transfer point for every pair of vehicles. We analyze underlying shortest-path problems and employ state-of-the-art routing algorithms to compute distances on-the-fly, which serves as the basis of dispatching requests with exact and up-to-date travel time information. We utilize constraints on existing routes, pruning techniques for transfer points, and both instruction- and thread-level parallelism to speed up the computation of the best assignment for every traveler. In addition to the exact variant, we propose a tunable heuristic approach that sacrifices solution quality in favor of improved running time. We evaluate our algorithm on a large road network with realistic input sets (up to 150000 requests). We demonstrate the effectiveness of our speedup techniques and the heuristic. We show first results on the benefits of transfers for taxi sharing on dense request sets, proving that our algorithm is well suited for the analysis of taxi sharing with transfers on large input instances. 2012 ACM Subject Classification Applied computing -> Transportation; Theory of computation -> Shortest paths; Information systems -> Geographic information systems
The Berth Allocation and Scheduling Problem (BASP) is a critical optimization challenge in maritime logistics, aiming to assign arriving vessels to berths efficiently, while adhering to practical constraints. Exploiting the connection of BASP with the Heterogeneous Vehicle Routing Problem with Time Windows (HVRPTW), we propose a mixed integer linear programming (MILP) formulation for a variant of BASP which is of utmost importance in real-world scenarios: the Dynamic Discrete Berth Allocation and Scheduling Problem with Time Windows (DDBASPTW). Consequently, inspired by the wealth of constructive and improvement heuristics for VRP, we design, implement and experimentally evaluate three constructive heuristics, Nearest Neighbour (NN), Insertion (INS), a quick-and-dirty variant of Insertion (qd-INS), as well as two improvement heuristics, Swap and Reinsert, taking into consideration both the online and the offline scenario with respect to vessel arrivals. Finally, we propose, implement and experimentally evaluate, custom-tailored variants for DDBASPTW of a single-solution metaheuristic, the Adaptive Large Neighborhood Search (ALNS), and of two population-based metaheuristics, the Genetic Algorithm (GA) and the Cuckoo Search Algorithm (CSA), which are aimed to solve the offline version of the problem. An extensive experimental evaluation compares these techniques against a generic state-of-the-art MILP solver. Results demonstrate that certain variants of INS not only are extremely fast and deliver competitive solutions, achieving a practical trade-off between execution times and quality of solutions. The improvement heuristics further refine the initial solutions, especially for weaker constructive approaches, offering a lightweight yet effective enhancement mechanism. The metaheuristics consistently yield high-quality solutions with significantly lower computational times compared to the exact MILP solver, making them well-suited for use in real-time or large-scale operational environments. 2012 ACM Subject Classification Theory of computation -> Theory and algorithms for application domains
In the sequential resource allocation problem there is a single divisible resource that is divided over a number of clients. Allocations are made in a predetermined order and only upon arrival at a client their demand for the resource is revealed; only the probability distribution of the demand of every client is known to the supplier. We consider this problem from a fairness perspective, where the aim is to balance allocations between individual clients. Several allocation policies have been proposed in the literature. In this work, we introduce a new, non-adaptive policy based on linear programming that can also incorporate group fairness. In addition, we provide an extensive computational study to compare allocation policies on several fairness measures. Using an optimized implementation of existing methods, we are able to evaluate significantly larger problem instances than those previously considered in the literature. 2012 ACM Subject Classification Applied computing -> Operations research
Multi-criteria route planning arises naturally in real-world navigation scenarios where users care about more than just one objective - such as minimizing travel time while also avoiding steep inclines or unpaved surfaces or toll routes. To capture the possible trade-offs between competing criteria, many algorithms compute the set of Pareto-optimal paths, which are paths that are not dominated by others with respect to the considered cost vectors. However, the number of Pareto-optimal paths can grow exponentially with the size of the input graph. This leads to significant computational overhead and results in large output sets that overwhelm users with too many alternatives. In this work, we present a technique based on the notion of regret minimization that efficiently filters the Pareto set during or after the search to a subset of specified size. Regret minimizing algorithms identify such a representative solution subset by considering how any possible user values any subset with respect to the objectives. We prove that regret-based filtering provides us with quality guarantees for the two main query types that are considered in the context of multi-criteria route planning, namely constrained shortest path queries and personalized path queries. Furthermore, we design a novel regret minimization algorithm that works for any number of criteria, is easy to implement and produces solutions with much smaller regret value than the most commonly used baseline algorithm. We carefully describe how to incorporate our regret minimization algorithm into existing route planning techniques to drastically reduce their running times and space consumption, while still returning paths that are close-to-optimal. 2012 ACM Subject Classification Theory of computation -> Data structures design and analysis; Theory of computation -> Shortest paths
We offer a geometric perspective on the problem of integrated periodic timetabling and passenger routing in public transport. Inside the space of periodic tensions, we single out those regions, where the same set of paths provides shortest passenger routes. This results in a polyhedral subdivision, which we combine with the known decomposition by polytropes. On each maximal region of the common refinement, the integrated problem is solvable in polynomial time. We transform these insights into a new geometry-driven primal heuristic, integrated tropical neighborhood search (ITNS). Computationally, we compare implementations of ITNS and the integrated (restricted) modulo network simplex algorithm on the TimPassLib benchmark set, and contribute better solutions in terms of total travel time for all but one of the twenty-five instances for which a proven optimal solution is not yet known. 2012 ACM Subject Classification Applied computing -> Transportation; Mathematics of computing -> Combinatorial optimization; Theory of computation -> Network optimization
We consider a scheduling environment in which jobs are associated with machine-dependent due-dates. This natural setting arises in systems where clients' tolerance depends on the service provider. The objective is to maximize throughput, defined as the number of non-tardy jobs. The problem exhibits significant differences from previously studied scheduling models. We analyze its computational complexity both in general and for the special case of unit-length jobs. In the unit-length setting, we provide an optimal algorithm that also extends to cases with machine-dependent release times and machine-dependent weights (i.e., rewards depending on the machine that completes the job). For jobs with different lengths, we show that even the unweighted problem without release times, with only two different lengths, specifically, for all j, pj is an element of {1, 2}, is APX-hard. To isolate the role of machine-dependent due-dates in this hardness result, we present an optimal algorithm for the case where all pj is an element of 1, 2 and due-dates are not machine-dependent. This algorithm further extends to instances with a constant number of integer processing times. 2012 ACM Subject Classification Theory of computation -> Scheduling algorithms
We introduce the line-based dial-a-ride problem with transfers (liDARPT), a variation of the wellstudied dial-a-ride problem (DARP), where vehicles transport requests on-demand but are constrained to operate along a set of lines, and passengers are allowed to transfer between lines on their journey. We develop an event-based solution approach for the liDARPT that relies on the construction of an event-based graph and uses a MILP to find optimal circulations in the event-based graph. To make this solution approach effective, we devise a pre-processing routine to limit the size of the event-based graph. We extensively test our approach on novel benchmark instances, inspired by real-life long-distance bus networks. In our experiments, problem instances with up to 80 requests can be solved to optimality within 15 minutes, and an average of 99.69% of requests are accepted in all instances solved to optimality. 2012 ACM Subject Classification Applied computing -> Transportation
We study the periodic assignment problem, in which a set of periodically repeating tasks must be assigned to workers within a repeating schedule. The classical efficiency objective is to minimize the number of workers required to operate the schedule. We propose a O(n log n) algorithm to solve this problem. Next, we formalize a notion of fairness among workers, and impose that each worker performs the same work over time. We analyze the resulting trade-off between efficiency and fairness, showing that the price of fairness is at most one extra worker, and that such a fair solution can always be found using the Nearest Neighbor heuristic. We characterize all instances that admit a solution that is both fair and efficient, and use this result to develop a O(n log n) exact algorithm for the fair periodic assignment problem. Finally, we show that allowing aperiodic schedules never reduces the price of fairness.
Network design problems have been studied from the 1950s, as they can be used in a wide range of real-world applications, e.g., design of communication and transportation networks. In classical network design problems, the objective is to minimize the cost of routing the demand flow through a graph. In this paper, we introduce a generalized version of such a problem, where the objective is to tradeoff routing costs and delivery speed; we introduce the concept of speed-coverage, which is defined as the number of unique items that can be sent to destinations in less than 1-day. Speedcoverage is a function of both the network design and the inventory stored at origin nodes, e.g., an item can be delivered in 1-day if it is in-stock at an origin that can reach a destination within 24 hours. Modeling inventory is inherently complex, since inventory coverage is described by an integer function with a large number of points (exponential to the number of origin sites), each one to be evaluated using historical data. To bypass this complexity, we first leverage a parametric optimization approach, which converts the non-linear joint routing and speed-coverage optimization problem into an equivalent mixed-integer linear program. Then, we propose a sampling strategy to avoid evaluating all the points of the speed-coverage function. The proposed method is evaluated on a series of numerical tests with representative scenarios and network sizes. We show that when considering the routing costs and monetary gains resulting from speed-coverage, our approach outperforms the baseline by 8.36% on average. 2012 ACM Subject Classification Applied computing -> Transportation
We study the complexity of the directed periodic temporal graph realization problem. This work is motivated by the design of periodic schedules in public transport with constraints on the quality of service. Namely, we require that the fastest path between (important) pairs of vertices is upper bounded by a specified maximum duration, encoded in an upper distance matrix D. While previous work has considered the undirected version of the problem, the application in public transport schedule design requires the flexibility to assign different departure times to the two directions of an edge. A problem instance can only be feasible if all values of the distance matrix are at least shortest path distances. However, the task of realizing exact fastest path distances in a periodic temporal graph is often too restrictive. Therefore, we introduce a minimum slack parameter k that describes a lower bound on the maximum allowed waiting time on each path. We concentrate on tree topologies and provide a full characterization of the complexity landscape with respect to the period Delta and the minimum slack parameter k, showing a sharp threshold between NP-complete cases and cases which are always realizable. We also provide hardness results for the special case of period Delta= 2 for general directed and undirected graphs. 2012 ACM Subject Classification Theory of computation -> Graph algorithms analysis; Mathematics of computing -> Discrete mathematics
Train timetables can be represented as event graphs, where events correspond to a train passing through a location at a certain point in time. A visual representation of an event graph is important for many applications such as dispatching and (the development of) dispatching software. A common way to represent event graphs are time-space diagrams. In such a diagram, key locations are visualized on the y-axis and time on the x-axis of a coordinate system. A train's movement is then represented as a connected sequence of line segments in this coordinate system. This visualization allows for an easy detection of infrastructure conflicts and safety distance violations. However, time-space diagrams are usually used only to depict event graphs that are restricted to corridors, where an obvious ordering of the locations exists. In this paper, we consider the visualization of general event graphs in time-space diagrams, where the challenge is to find an ordering of the locations that produces readable drawings. We argue that this means to minimize the number of turns, i.e., the total number of changes in y-direction. To this end, we establish a connection between this problem and Maximum Betweenness. Then we develop a preprocessing strategy to reduce the instance size. We also propose a parameterized algorithm and integer linear programming formulations. We experimentally evaluate the preprocessing strategy and the integer programming formulations on a real-world dataset. Our best algorithm solves every instance in the dataset in less than a second. This suggests that turn-optimal time-space diagrams can be computed in real time. 2012 ACM Subject Classification Theory of computation -> Graph algorithms analysis; Theory of computation -> Fixed parameter tractability; Theory of computation -> Computational geometry
The Multi-Capacity Fixed-Charge Network Flow (MC-FCNF) problem, a generalization of the Fixed-Charge Network Flow problem, aims to assign capacities to edges in a flow network such that a target amount of flow can be hosted at minimum cost. The cost model for both problems dictates that the fixed cost of an edge is incurred for any non-zero amount of flow hosted by that edge. This problem naturally arises in many areas including infrastructure design, transportation, telecommunications, and supply chain management. The MC-FCNF problem is NP-Hard, so solving large instances using exact techniques is impractical. This paper presents a genetic algorithm designed to quickly find high-quality flow solutions to the MC-FCNF problem. The genetic algorithm uses a novel solution representation scheme that eliminates the need to repair invalid flow solutions, which is an issue common to many other genetic algorithms for the MC-FCNF problem. The genetic algorithm's efficiency is displayed with an evaluation using real-world CO2 capture and storage infrastructure design data. The evaluation results highlight the genetic algorithm's potential for solving large-scale network design problems.
We study the Capacitated Team Orienteering Problem, where a fleet of vehicles with capacities have to meet customers with known demands and prizes for a single commodity. The objective is to maximize the total prize and to assign a sequence of customers to each vehicle while keeping the total distance traveled within a given budget and such that the total demand served by each vehicle does not exceed its capacity. The problem has been widely studied both from a theoretical and a practical point of view. The contribution of this paper is twofold: (1) We advance the theoretical knowledge on the problem by providing new approximation algorithms that achieve, under some natural assumption, improved approximation ratios compared to the current best algorithms; (2) We propose four efficient heuristics that outperform the current state-of-the-art practical methods in the sense that they compute solutions that collect nearly the same prize in a significantly smaller running time. We also experimentally test the scalability of the new heuristics, showing that their running time increases approximately linearly with the size of the input, allowing us to process large graphs which were not possible to analyze before.
Simulation of urban rail networks provides useful information to optimize traffic management strategies w.r.t. goals such as satisfaction of passenger demands, adherence to schedules or energy saving. Many network models are too precise for the analysis needs, and do not exploit concurrency. This results in an explosion in the size of models, and long simulation times. This paper presents an extension of Petri nets that handles trajectories of trains, passenger flows, and scenarios for passenger arrivals. We then define a fast event-based simulation scheme. We test our model on a real case study, the Metro of Montreal, and show that full days of train operations with passengers can be simulated in a few seconds, allowing analysis of quantitative properties.
