· · · Institutional Repository

Suboptimaal weggebruik leidt bij hogere intensiteiten tot een slechtere benutting van de beschikbare infrastructuur en daarmee mogelijk tot files. In dit onderzoek is geanalyseerd of het geven van individuele aanwijzingen aan automobilisten een positief effect heeft op de verkeersdoorstroming en de stabiliteit. Voor twee verkeerssituaties, namelijk het eerder anticiperen op langzamere voertuigen en verbeteren van het invoegproces, is door simulaties met Fosim aangetoond dat het geven van de individuele aanwijzingen tot een verbeterde doorstroming en een stabielere verkeersstroom leidt.

Like in many countries congestion is also a big problem in the Netherlands. Congestion is damaging for the economy, the environment, and our own health. There is a big need for solving this problem. Possible solutions are: increasing road capacity, road pricing, fuel taxation, using other modes of transport, etc. Road pricing is seen as one of the best ways to decrease congestion. The traffic authorities want to influence travellers' behaviour with the help of road pricing causing them to change route, change departure time, change mode of travelling, etc., resulting in a more efficient use of transportation resources and relieving congestion on the roads. The road pricing problem can be seen as a situation where two decision makers have conflicting interests. The traffic authorities want to find an optimal toll pattern to achieve a situation in which the road system is in optimal form. I n other words, the traffic authorities want to minimise the total time travellers spend on the network. Travellers, on the other hand, want to minimise their own travel times. It is possible that there are travellers in the optimal network conditions, that can improve their own travel time. If they indeed choose another route for improving their travel time, network conditions deteriorate. The road pricing problem is restricted by some boundary conditions and basic assumptions of which 'the traffic assignment and the traffic flows will be static' is the most important. This research will focus on solving a mathematical representation of the road pricing problem using different methods taking into account the complexity and the properties of the road pricing model. Restrictions are made about the value of toll (lower and upper bound) and the number of roads that may be tolled (not all roads may/can be tolled). Therefore, the purpose of this thesis is not to toll every link, but to achieve a system state that is near the system optimum by charging toll on a subset of links and taking into account a lower and an upper bound for the tolls. This means that i t is a constrained optimisation problem and that some of the variables of the solution may be lying at the boundaries of their allowed ranges. The plan of approach for this research is to do first a literature study on bi-level optimisation and later on solution methods. The road pricing problem can be written mathematically as a two-level optimisation model (a special case of bi-level optimisation). The mathematical model consists of an upper and a lower level. The lower level is a representation of the behaviour of the travellers and in the upper level the traffic authorities are modelled. Both levels consist of an objective function and some constraints. The complexity and the convexity of the road pricing model are discussed. After that some solution methods and their convergence are discussed. Finally, these methods are tested on four networks and the results are reviewed. In complexity theory there are four classes of problems: P, NP, NP-complete, and NP-hard. Problems belonging to the class P are called (computationally) tractable. Problems that belong to the other three classes are called intractable problems. In the literature much is proved about the complexity of a bi-level problem. The conclusion is that the road pricing problem is NP-hard. This means that it is not likely to find an exact solution with a polynomial-time algorithm. Heuristics will therefore be used to solve the road pricing model. It is plausible that the road pricing model is a convex optimisation problem. In that case existing solution methods in order to find local optima can be used. Therefore, line search, first order gradient, direction set, and simplex methods are used to solve the road pricing model. These methods are implemented and eleven test cases are considered to assess and compare the methods. The seven best performing methods in the first four test cases are tested on the resulting test cases. These are the EDO/BS, the SD/BS, the Hestenes/Stiefel, the cyclic coordinate, the Rosenbrock, the Powell I I , and the complex method. They are compared per test case with respect to the value of the objective function (1), the number of the function evaluations (2), and the number of iterations (3). The methods are given notes from 1 t i l l 7 for these three criteria. The best performing method is given number 7 and the worst is given number 1. The average over all the test cases is shown in Table 6.1. Best performing means the smallest value of the objective function, the smallest number of function evaluations, or the smallest number of iterations. This is not the best way for comparing the methods, but it forms a notion of the performance of the methods compared with each other.

Road pricing is a policy measure under consideration by many goverments and road authorities. Although objectives may be different any road pricing measure will impact the behaviour of travellers and the flow of traffic. In this research we specifically looked at the effects of road pricing measures where the prices are differentiated by time and place. Different choice models were estimated using data from a stated preference survey that was designed within this research. We find that the sensitivities of commuters towards changing departure time exist for both departure and arrival time, they are non-linear and they depend on personal constraints. Also the unreliability in travel time affects the departure time choice of commuters. Since the sensitivities towards rescheduling are high, it is to be recommended to introduce supporting policy measure that lower sensitivities towards rescheduling. In this research we also developed a dynamic modelling framework that can forecast the traffic effects of time-varying road pricing measures on large networks. A dynamic, multi-user class, equilibrium model is used in conjunction with a departure time choice, and an elastic demand model. The departure time choice models were estimated in this research. The model framework was applied on a reward policy case study and results show that traffic flow changes depend on both reward and participation levels in a non-linear manner. In order to assess the potential effects of time-varying road pricing measures it is therefore necessary to apply dynamic models since only those models can appropriately describe changes in congestion levels on a network.

This research focuses on the influence of reliability of urban public transport services on network evaluation and network design. Public transport services suffer from all kind of disturbances, varying demand patterns, weather conditions, incidents, road works, et cetera. Current practice uses all kinds of operational measures to ensure the quality of the services despite of these disturbances. In network evaluation and network design, however, reliability is usually not considered. The research presents an extensive overview of phenomena affecting public transport service quality, as well as a framework and instrument to determine the influence of the network on reliability indicators and the total network performance. Using case studies a number of measures is analysed with respect to their effectiveness. This research demonstrates that already at the level of public transport service network design there are possibilities to improve service reliability: for example networks in which travellers have multiple options to travel to their destination and networks in which line lengths are limited. However, it is dependent on the specific situation whether this will lead to an improvement of the overall network performance as well. For rail-bound public transportation, for instance tram, is proves to be possible to implement smart infrastructure measures, for instance turning facilities and shortcuts, to improve both service reliability and total network performance.

Incidents cause a large part of the delays in road networks. This is caused by a decrease of the capacity at the incident site. A detailed knowledge of the queue discharge rate can improve for instance the traffic prediction and thereby improve delay information or routing advice. Therefore, this study determines the queue discharge rate for many incident locations during an incident situation and these are compared with the queue discharge rate at the same location in normal conditions. Ninety incidents meet the requirements to apply the proposed methodology. It is found in case a driving lane is blocked, the queue discharge rate for each available lane is reduced by 50%. In case the driving lanes are open but there is a distraction of an incident at the emergency lane or on the roadway for the opposite direction, the queue discharge rate is reduced by 30%.

Instructing evacuees on their departure time, destination, and route can lead to more efficient evacuation traffic operations. While current evacuation plan optimization techniques are limited to assessing mandatory evacuation where travelers strictly follow the instructions, in reality a share of travelers likely decides not to comply. Here we show how 1) traveler compliance behavior affects evacuation efficiency, and 2) evacuation efficiency can be improved in case of less than full compliance when this traveler compliance is anticipated on. To this end, we use heuristic ant colony optimization in combination with the evacuation network model EVAQ in which compliance behavior is explicitly accounted for. The method is described and illustrated using the case study describing the evacuation of the Walcheren peninsula, the Netherlands. The method and application underline the conclusion that traveler compliance is an essential consideration while deciding on the appropriate evacuation instructions to be given. Also, the approach proposed here gives direction to further research along this line contributing to the understanding of the impact of traveler compliance behavior and its assessment in evacuation planning.

Stochastic supply and fluctuating travel demand lead to stochasticity in travel times and travel costs experienced by travelers from time to time within a day and at the same time from day to day. Many studies show that travel time un-reliability has significant impacts on traveler’s choice behavior under uncertainty. Different theories are applicable to describe and model traveler’s choice behavior under uncertainty. This paper presents several behavioral models based on differ-ent behavioral hypotheses under uncertainty in the framework of utility theory. An analytical approach is employed to investigate the generality and the relationships among the different behavioral models for modeling traveler’s route/departure time choice behaviors under uncertainty. This paper shows the principal equiva-lence of different behavioral models and the condition at which the equivalence is maintained. The parameters of the utility components are explored with a theoreti-cal approach and a simulation-based approach. This paper explores which utility components should be incorporated in the behavioral model, which has substantial meaning for the researchers and practitioners to gain insights into the relationships of different behavioral models for modeling traveler’s choice behavior under un-certainty. It is concluded that the mean variance approach is a special case of the scheduling approach and of the generalized utility approach. The generalized utili-ty approach is equivalent to the scheduling approach under a certain condition.

Zoetermeer in The Netherlands. During peak hours the frequency on some trajectories is about 24 vehicles an hour. Dealing with these high frequencies and offering travelers a high quality product, according to waiting times as well as the probability of getting a seat, the operator designed a three step controlling philosophy. The first step is to prevent deviations to occur: the infrastructure is exclusive right of way as much as possible and at intersections RandstadRail gets priority over the other traffic. RandstadRail stops at every stop and never leaves before the scheduled time. The second step in the philosophy is dealing with deviations by planning extra time in the schedule at stops, trajectories and terminals. Small deviations can be solved in this way. The final step to get vehicles back on schedule is done by the traffic control centre: they have a total overview of all vehicles and they can respond to disturbances like slowing down vehicles nearby a delayed vehicle. Experiencing major disturbances rerouting and shortening of lines is possible. RandstadRail is in operation since 2007. The actual data of the performance is used to analyze the actual effects of the control philosophy. It is shown that due to the applied measures the variability of the driving times is reduced. Punctuality has increased as well. This leads to a higher level of service, creating shorter travel times, a better distribution of passengers over the vehicles.

Stability and robustness of a railway timetable are essential properties for punctual and reliable operations. Timetable performance evaluation is therefore an important aspect in the timetable design process. In particular, the stability and recoverability properties of a timetable with respect to daily process time variations must be well analysed. The timetable must be able to recover from primary delays due to stochastic process times and it must be robust against secondary delays due to train interactions. This paper presents a stability analysis approach based on stochastic max-plus linear system theory. Stochastic counterparts of well-established concepts from the deterministic max-plus stability analysis are proposed, like timetable stability and realizability. General probability distributions can be used to model the primary stochastic behaviour of process times, while delay propagation due to timetable and infrastructure constraints are computed from the stochastic recursive system equations. Recently developed powerful algorithms can be utilized to analyse and improve large-scale stochastic systems, and to establish the amount of stochastic variations that a timetable can absorb without external control.

Railway traffic experiences disturbances during operations while dispatchers need actions to restore feasibility and limit spreading of delays through the network. To help the dispatcher in such task, the dispatching support tool ROMA (Railway traffic Optimization by Means of Alternative graphs) has been designed and implemented. We report on enhancements to the underlaying train dispatching model as well as to the solving algorithms studied in order to tackle the increased complexity of busy stations with multiple conflicting paths and high service frequency. The system is compared with straightforward rules and the current approach in the Netherlands. Extensive computational studies based on accepted statistical distributions of delays assess the effectiveness of the ROMA tool.

The paper discusses the current state of research concerning railway network timetabling and traffic management. Timetable effectiveness is governed by frequency, regularity, accurate running, recovery and layover times, as well as minimal headway, buffer times and waiting times. Analytic (queuing) models and stochastic micro-simulation are predominantly used for estimation of waiting times and capacity consumption anlong corridors and in stations, while combinatorial models and stability analysis are suitable for network timetable optimisation. Efficient traffic management can be achieved by real-time monitoring, fusion, analysis and rescheduling of railway traffic in case of disturbances. Real-time simulation, optimisation and impact evaluation of dispatching measures can improve the effectiveness of rescheduling and traffic management. The display of dynamic signal and track occupancy data in driver cabins, as RouteLint developed by ProRail, can support anticipative actions of the driver in order to reduce knock-on delays and increase throughput.

Charging road users for ‘bad behaviour’ may be used to improve mo-bility and alleviate congestion. Rewarding road users for ‘good behaviour’ may be used to achieve the same objective, however it sends out a positive instead of a negative incentive, and will therefore be more easily accepted by travellers con-fronted with such pricing policies. In the Netherlands, experiments have been conducted in order to investigate travel behaviour when people are rewarded for avoiding the peak period by car. Furthermore, some real life rewarding projects were defined in order to positively influence traffic conditions in situations where road works may temporarily worsen the traffic conditions. This paper presents outcomes of these studies. It is found that in all four case studies the (volunteer-ing) travellers are willing to adjust their departure time significantly for some modest monetary rewards. Also route and mode shifts were observed. The effec-tiveness of the rewarding scheme is strongly determined by the setup and the loca-tion (definition of peak period, reward level, available public transport, available alternative routes, etc.).

De capaciteitsverdeling van het spoor aan de diverse vervoerders gebeurt via basisuurpatronen (BUP’s) zoals vastgelegd in de Netverklaring door ProRail. De TU Delft heeft de analysetool PETER ontwikkeld waarmee de stabiliteit van basisuurpatronen op netwerkniveau kunnen worden geanalyseerd op een transparante, non-discriminatoire, controleerbare en reproduceerbare manier. PETER (Performance Evaluation of Timed Events in Railways) gaat uit van de basisuurpatronen voor alle treinseries op het hoofdnetwerk en regionale lijnen inclusief reizigersaansluitingen, materieelkeringen en infrastructuurbeperkingen zoals opvolg- en overkruistijden. Door deze stabiliteitsanalyse mee te nemen in het capaciteitsverdelingsproces kan de stabiliteit van basisuurpatronen worden getoetst met betrekking tot netwerkafhankelijkheden en bijvoorbeeld het uitbuigen van dienstregelingspaden in het dienstregelingsontwerpproces. Aan de hand van een case-studie voor het basisuurpatroon 2007 wordt de methodiek toegelicht.

Process operators at large railway stations have the difficult task to secure a fluent train traffic flow while minimizing deviations from established timetables. Variation in actual train departure times is inevitable due to many circumstances such as arrival delays and fluctuations in alighting and boarding time, even if some buffer time is contained in the dwell time. Moreover, a departure may be delayed by waiting for a feeder train to secure a connection and by conflicting train movements prohibiting an outbound train path. The predictability of train processes is even more degraded when in similar situations different control actions are pursued depending on for instance individual dispatchers. In the Netherlands, passenger train services operate basically according to a cyclic timetable, repeating the same arrival and departure times each hour, with the exception of additional passenger trains in rush hours and freight trains that are scheduled in between the regular train services. It is hence anticipated that the traffic processes are mainly variations on a repetitious pattern. Analysis of historical realization data then yields operational insight that can be used to improve or support process management. To gain accurate operations data a software tool, TNV-Prepare, has been developed that filters relevant train detection data from train describer records. This paper starts with a brief account of the collection and preparation of train detection data. Then for the particular case of station Eindhoven a detailed punctuality analysis is reported including the performance of dwell and transfer times (tightness or possible recovery time) and train waiting times to secure connections. Departure delays are predicted from arrival delays using regression analysis, whereas the remaining noise is attributed to human factors.

The estimation of the precise arrival and departure times of trains at stations is done by means of a software tool that extracts the occupation and clearance times of each train per track section of the Dutch Railways‘ network. The software tool was applied to the whole automatically collected set of signal data of the area of Eindhoven during one week in September 1997 comprising 140 MB. A total of 1846 trains were detected and used for further analysis. The detailed statistical analysis of the distribution of the train arrivals, dwell times and departures shows a systematic mean arrival delay of almost every line (InterCity, InterRegio, AggloRegio) ranging up to 138 sec per train. All of the IC-lines except the line starting nearby at Heerlen showed a punctuality of less than 90% when punctuality is defined as having less than 3 minutes of delay. The IR- and AR-lines arrived more punctual. The dwell times of all the lines lasted on the average 30 to 90 sec longer than scheduled and more than two thirds of the trains arriving late extended the planned dwell time. Consequently, the mean departure delays per line ranged between 1.5 and 2.5 minutes. There is much statistical evidence that the arrival delays and the dwell times fit to normality, whereas the departure delays have a clear negative exponential distribution.

Providing travel time information to travelers on available route alternatives in traffic networks is widely believed to yield positive effects on individual drive behavior and (route/departure time) choice behavior, as well as on collective traffic operations in terms of, for example, overall time savings and—if nothing else—on the reliability of travel times. As such, there is an increasing need for fast and reliable online travel time prediction models. Previous research showed that data-driven approaches such as the state-space neural network (SSNN) are reliable and accurate travel time predictors for freeway routes, which can be used to provide predictive travel time information on, for example, variable message sign panels. In an operational context, the adaptivity of such models is a crucial property. Since travel times are available (and, hence, can be measured) for realized trips only, adapting the parameters (weights) of a data-driven travel time prediction model such as the SSNN is particularly challenging. This paper proposes a new extended Kalman filter (EKF) based online-learning approach, i.e., the online-censored EKF method, which can be applied online and offers improvements over a delayed approach in which learning takes place only as realized travel times are available.

Railway is an important and sustainable transportation mode, which despite good potentials results in a limited attractiveness, mostly due to the perceived consequences of unreliability. In fact, busy railway networks with frequent and heterogeneous services are highly sensitive to delay propagation, due to conflicts along lines and at stations. Primary delays due to accidents or failures spread easily in the network as consecutive delays, affecting passengers. Buffer times are able to reduce impact of delays, but dispatching actions are required in real-time to solve infeasible train traffic situations and limit delay propagation. Dispatchers currently limit their actions to strongly disturbed situations, and determine their decision combining their experience with the limited available information on the traffic status, often with suboptimal outcomes. Much better performance could be achieved if proactive and informed decisions were taken, based on detailed traffic information, extended for a period of traffic prediction, and based on advanced mathematical methods to deliver locally and globally optimal dispatching actions. This thesis investigates real-time railway traffic management, i.e. optimizing train orders, routes, departure times in real-time based on the actual, unpredictable situation of train traffic. Support to dispatchers is most required for difficult problems due to geographical size, traffic frequency and density, amount of traffic, operational policies adopted. A flexible and detailed mathematical formulation based on alternative graphs models precisely railway operations at the level of block sections to microscopically check feasibility of train movements along open track and in complex and busy interlocking areas according to the blocking time theory. Evolution of train traffic is simulated precisely based on the current situation of train positions, speeds, traffic and infrastructure status, and considering operational constraints due e.g. to crew or rolling stock plans, transfer connections. A variety of powerful advanced algorithms compute retiming, reordering and rerouting actions to solve potential conflicts and optimally minimize delay propagation, considering the available capacity at stations and lines in microscopic detail. Very large areas are decomposed in smaller, local dispatching problems that are solved independently, while an intelligent coordination procedure harmonizes their solutions and guides the solvers toward locally feasible and globally optimal solutions. The model is also able to take into account interesting objectives and requirements from stakeholders, such as energy-efficient operations, feasible emergency timetables for exceptional situations and heavily disrupted operations, online management of passenger connections. The investigation on comprehensive mathematical models, algorithms and approaches resulted in a laboratory dispatching decision support system, able to optimize and report a quantitative evaluation of advanced dispatching solutions of proven feasibility. Through real-time railway traffic management, better and more attractive railway services can be delivered to passengers with benefits for the whole society.

The dissertation research deals with the design of the controller for a Dynamic Route Guidance System (DRGS) that uses both road side variable message signs as well as in car personal navigation devices. The DRGS is able to determine user- and system optimal route guidance advices in real-time for realistic networks using a model predictive control approach. A case study is performed for the traffic network of Rotterdam where the benefit of this kind of route guidance to reduce total delay are illustrated.

In highly utilized rail networks, as in the Netherlands, conflicts and subsequent train delays propagate considerably in time and space during operations. In order to realistically forecast and minimize delay propagation, there is a need to extend short-term traffic planning up to several hours. On the other hand, as the magnitude of the time horizon increases the problem becomes computationally intractable and hard to tackle. In this paper, we decompose a long time horizon into tractable intervals to be solved in cascade with the objective of improving punctuality. We use the ROMA dispatching system to pro-actively detect and globally solve conflicts on each time interval. The future evolution of railway traffic is predicted on the basis of the actual track occupation, theDutch signaling system and dynamic train characteristics. Extensive computational tests are carried out on the railway dispatching area between Utrecht and Den Bosch.