Neues verkehrswissenschaftliches Journal - Ausgabe 16 E-Book

Ullrich Martin

• Herausgeber: Books on Demand
• Kategorie: Fachliteratur
• Sprache: Englisch

Seitenzahl: 203

Veröffentlichungsjahr: 2017

Beschreibung

In urban mixed traffic zones, urban rail-bound transport interacts with road traffic, which includes motorized road traffic as well as non-motorized road traffic. The quality of operation and the capacity of urban rail-bound transport are heavily influenced by the road traffic in urban mixed traffic zones. Capacity research is one of the most important methods for the validation of timetables on an existing or planned infrastructure, for adequate railway capacity design of the infrastructure, as well as the performance evaluation for railway operation. For the urban mixed traffic zones, two developed approaches in this research can be used to carry out capacity research with consideration of the road traffic influences on urban rail-bound transport. As a result of the stochastic influences of road traffic on urban rail-bound transport, the waiting time function will be adapted to mixed traffic conditions, which can derive the recommended area of traffic flow for the situation of urban mixed traffic more plausibly. In addition, an algorithm for modeling an event-driven system is developed for preliminary determining the results of capacity research. Accordingly, the operating performance for urban rail-bound transport with road traffic influences can be determined and evaluated.

Leseprobe

Preface

In 2013, the research project with the topic “Capacity Research in Urban Rail-Bound Transportation with Special Consideration of Mixed Traffic” was proposed and granted by the Deutsche Forschungsgemeinschaft (DFG). The research project was completed in 2015, in which the application of capacity research for the evaluation of rail-bound systems is applied to urban rail-bound transport with external influences caused by road traffic in urban mixed traffic zones.

In this completed research project, a simulation model for urban rail-bound transport with consideration of road traffic influences can be built through two developed approaches. The modeling approach—through modeling the road traffic as urban rail-bound transport in a simulation model, and the distribution approach—using mathematical distributions to convert the road traffic influences to perturbations in the simulation model, the influences of road traffic on urban rail-bound transport can be determined for capacity research. One important intermediate result of capacity research, the model function of the waiting time function, is further improved to be adapted to the urban rail-bound system with consideration of external influence (caused by road traffic in urban mixed traffic zones). In addition, a developed algorithm with an event-driven system for urban mixed traffic can be used for the determination of the waiting time function and the throughput capacity for the evaluation of urban rail-bound transport with consideration of mixed traffic, which is also applied for the preliminary assessment of the significances of road traffic influences at each mixed traffic zone in a whole investigated area.

With the results of this project, the method and application of capacity research for the macroscopic and mesoscopic evaluation is supplied and expanded to include urban mixed traffic, so that the quality of operation and the operational performance of urban rail-bound transport with road traffic influences at mixed traffic zone can be further evaluated.

Stuttgart, December 2016

Ullrich Martin

Table of Content

Table of Content

List of Figures

List of Tables

1 Introduction

2 Capacity Research of Rail-Bound Transport

2.1 Overview

2.2 Relevant Terms

2.3 Methods of Capacity Research

2.3.1 Analytical Method

2.3.2 Simulation Method

3 External Influences on Urban Rail-Bound Transport in Urban Mixed Traffic Zone

3.1 Urban Mixed Traffic

3.1.1 Overview

3.1.2 Related Findings to Road Traffic

3.2 Conditions of Investigated Mixed Traffic Zone

3.2.1 Traffic Control Signaling System in Mixed Traffic Zones

3.2.2 Special Conditions

3.3 Influential Parameters

3.4 Sensitivity Analysis

3.4.1 Existing Methods of Sensitivity Analyses

3.4.2 Sensitivity Analysis of Qualitative Influential Parameters

4 Modeling Approach for Road Traffic Influences on Urban Rail-Bound Transport

4.1 Basic Concept

4.2 Methodology for Modeling Approach

4.2.1 Simulation Model

4.2.2 Capacity Research of Modeling Approach

4.3 Applications of Modeling Approach

5 Distribution Approach for the Description of Road Traffic Influences on Urban Rail-bound Transport

5.1 Basic Concept

5.2 Perturbation Parameters for Mixed Traffic Zones

5.2.1 Definition of Perturbation Parameters

5.2.2 Basic Calibration Algorithm

5.3 Methodology of the Distribution Approach

5.3.1 Overview

5.3.2 Procedure of the Distribution Approach

5.4 Capacity Research with the Distribution Approach

6 Derivation of an Adapted Waiting Time Function

6.1 Overview

6.2 Existing Waiting Time Function

6.2.1 Basic Concept

6.2.2 Waiting Time Function by Ludwig and Hertel

6.2.3 Waiting Time Function by Chu

6.3 Adapted Waiting Time Function for Urban Mixed Traffic

6.3.1 Adapted Waiting Time Function

6.3.2 Fit Options

6.4 The Algorithm for the Derivation of the Adapted Waiting Time Function with Road Traffic Influences

6.4.1 Basic Concept

6.4.2 Algorithm Description

6.4.3 Determination of the Parameters

6.4.4 Event-driven Cases

6.4.5 Algorithm for a Whole Investigated Area

7 Comparison of the Approaches and Validation of the Results

7.1 Analysis of the Capacity Research Results with Various Approaches

7.2 Applicable Conditions

8 Summary

Abbreviations

Formula Symbols

Glossary

Bibliography

Appendix

List of Figures

Figure 2-1: Evaluation of Performance for Capacity Research (Source: [Chu 2014] and [Li 2015])

Figure 3-1: Two Conditions of Urban Rail-bound Transport with Road Traffic Influences Discussed in this Research Project

Figure 3-1: Probability Distribution of Urban Rail-bound Transport Waiting Time with Road Traffic influences at Different Investigated Time Periods (Source: modified based on [Martin & Di Liu 2016])

Figure 3-1: Probability Distribution of Urban Rail-bound Transport Waiting time with Road Traffic Influences on Different Day Types (Source: modified based on [Martin & Di Liu 2016])

Figure 3-1: Infrastructure of an Investigated Example of Mixed Traffic Zone of Level Crossing for the Simulation Model

Figure 3-1: Results of Modeling Approach Comparison to that without Road Traffic Influences for an investigated Example (Source: modified based on [Martin & Di Liu 2016])

Figure 3-1: Procedure for Determination of the Perturbation Parameters PROT (Source: modified based on [Martin & Di Liu 2016])

Figure 3-1: Workflow of Capacity Research with the Distribution Approach

Figure 3-1: Results of the Distribution Approach Comparison to that without Road Traffic Influences for an Investigated Example (Source: modified based on [Martin & Di Liu 2016])

Figure 3-1: Relationship of the Average Number of Trains (Requests) between the Transient Phase and the Stationary Phase (Source: [Chu 2014])

Figure 3-1: Fitting Curve with the Waiting Time Function (6-7) with a Constant Parameter d3

Figure 3-1: Fitting Curve with the Waiting Time Function (6-10) with an additional Term d4/(1 – η)

Figure 3-1: Fitted Curve with the Waiting Time Function (6-11) with an additional Term d5/(1 – η)2/3

Figure 3-1: Comparison of the Fitted Curve of the Waiting Time Function (6-11) with and without Robust Regression of “Bisquare” Weights

Figure 6-6: Scheme of the Modeling of Event-Driven System based on the Algorithm

Figure 3-1: Microscopic Description of Routes at a Mixed Traffic Zone

Figure 3-1: An Example of Exclusive Routes Combination and Compatible Routes Combination in the investigated Example

Figure 3-1: The Effective Occupation Time of a Road Traffic Route with Different Phases of a Traffic Control Signal

Figure 3-1: Relationship between potential Occupation Time for Road Traffic (OROT) and the Hindrance Time of the Urban Rail-bound Transport by the Road Traffic (HURT) (Source: modified based on [Martin & Di Liu 2016])

Figure 3-1: The Workflow of the Developed Algorithm with Event-Driven Simulation

Figure 3-1: Event Trigger Conditions of Six Cases for Mixed Traffic Zone of Level Crossing

List of Tables

Table 4-1: Partly Scheduled Timetable of an Investigated Example of Mixed Traffic Zone of Level Crossing for Simulation Model

Table 6-1: Comparison of the Coefficient of Determination of the Adapted Waiting Time Function (6-7) and the Existing Waiting Time Function(6-5)

Table 6-2: Comparison of Coefficient of Determination of the Waiting Time Function (6-10) and the Existing Waiting Time Function(6-5)

Table 6-3: Comparison of the Coefficient of Determination of the Adapted Waiting Time Functions with the Four Parameters (6-10) and (6-11)

Table 6-4: The Locking Table of Route Related Locking for Urban Mixed Traffic of Example Mixed Traffic Zone (Source: modified based on [Martin & Di Liu 2016])

Table 6-5: Time Duration of Traffic Light Phases of Road Traffic Routes for the Investigated Example of Level Crossing

Table 6-6: Results of the Cases with and without Road Traffic Influences with the Two Approaches

1Introduction

Nowadays, the possibility for the construction of new railway infrastructure and the extension of existing infrastructure has declined after decades of development. However, economic development and an accompanying increase in the number of vehicles on the roads, especially in urban areas, has been observed. Important questions for the future research of urban mixed traffic include how to use the existing infrastructure as intensively and as reasonably as possible, while maximizing the capacity of the infrastructure without reducing the required operational quality. A key ingredient in doing this is by analyzing the waiting time function that reflects the relationship between the quality of operation (waiting time) and the capacity. The waiting time function can be derived from the results of simulations of stochastic influenced timetables with stepwise-varied traffic flows of trains (different compaction percentages of the given operating program) due to the operational hindrances between trains.

In urban mixed traffic zones, urban rail-bound transport (URT) interacts with road traffic (ROT), which includes motorized road traffic as well as non-motorized road traffic. Due to the inherent complexity of urban mixed traffic zones, existing studies in the field of capacity research have rarely considered the external influences caused by road traffic. These influences that impact the simulations lead to a deviation of the existing waiting time function. Hence, the results of capacity research without consideration of road traffic in mixed traffic zones are inaccurate. In order to improve the accuracy of capacity research, algorithms for capacity research of urban rail-bound transport were developed in this research project with special consideration given to the influences of road traffic in mixed traffic zones.

In this research project supported by the Deutsche Forschungsgemeinschaft with the reference number of “MA 2326/13-1”, methods for the capacity research of urban rail-bound transport were investigated with special consideration of the road traffic influences in mixed traffic zones. A general introduction to capacity research and two main methods of capacity research for rail-bound transport are presented in Chapter 2. To model the urban rail-bund transport with road traffic influences to a sufficient accuracy level without redundancy, the significant influential parameters were identified and studied with a sensitivity analysis in Chapter 3, which were based on data collected from on-site measurements. Two approaches (the modeling approach and the distribution approach) were developed to execute the capacity research with consideration of road traffic influences. With regards to the modeling approach, the road traffic was modeled comparable to trains in simulation tools to reflect the influences of road traffic, as described in Chapter 4. With regards to the distribution approach, the influences of road traffic on the urban rail-bound transport were modeled as operational perturbations in the simulation process with the help of a simulation tool, as described in Chapter 5. The capacity research was carried out with the simulation method assisted with the two approaches. Because of the additional external influences of road traffic, the existing waiting time function cannot fit the simulation results properly. An adapted waiting time function for capacity research with consideration of road traffic influences was further improved and is described in Chapter 6. Furthermore, an algorithm for determining the adapted waiting time function for capacity research was developed. Finally, the results of capacity research of the two approaches were compared and evaluated in Chapter 7.

2Capacity Research of Rail-Bound Transport

In this chapter, a general summary is made about capacity research for the rail-bound transport of the railway system. Firstly, Subchapter 2.1 manifests the basic concept of capacity research for the railway system. The relevant basic terms will be introduced in Subchapter 2.2. Two major methodologies for capacity research of rail-bound transport, which are the analytical method and the simulation method, will be presented in Subchapter 2.3.

2.1Overview

Capacity research is one of the most important methods for the validation of timetables on an existing or planned infrastructure, for adequate railway capacity design of the infrastructure, as well as the performance evaluation for railway operation [Pachl 2014]. In railway operation, capacity research is a significant issue that evaluates the operating performance of the railway systems. The operating performance is generally considered to be the quality of operation and the number of trains per time unit (hour) running in an investigated railway infrastructure network. The quality of operation usually refers to the waiting time (operational delays) in operation. Therefore, the capacity is inversely proportional to the quality of operation. One goal of capacity research is to improve the performance through optimizing the existing infrastructure and the operating program. Determination of the quality of operation for an investigated area with a concrete timetable and determination of recommended area of traffic flow with a rough operating program can both be achieved with railway capacity research.

2.2Relevant Terms

Relevant terms of capacity research for railway systems in this research project are defined as following:

2.3Methods of Capacity Research

There are various methods of capacity research, for the description of the operational quality and the further operating performance, the waiting time is one of the most important parameters of capacity research. Moreover, an intermediate result of capacity research is the throughput capacity for a specific operating program on a given railway infrastructure and during a specific time period. There are basically two methods to determine the waiting time and the throughput capacity for capacity research:

Analytical method

Simulation method

2.3.1Analytical Method

The analytical method is a common method for conducting capacity research and is based on mathematic analysis. Using this method allows for the calculation of the capacity of the railway infrastructure including the railway lines and railway nodes (railway stations), and waiting time by means of mathematical expressions given the infrastructure and the characteristics of the operating program ([Potthoff 1969], [Schwanhäußer 1978], [Kontaxi & Ricci 2010] and [Pachl 2014]). The infrastructure and timetable (operating program) are modeled with suitable mathematic models, such as parallel servers and the probability distribution. For the analytical method, the basic theories are the queuing theory and the probability theory [Potthoff 1972]. The investigated area of the railway system can be modeled as a corresponding queuing system. Accordingly, the waiting time can be derived by using mathematical models.

For the analytical method, the railway operation process can be described mathematically either by the deterministic expression or the stochastic expression. The analytical method is based on certain assumptions and an investigated operating program by using mathematical methods to assess the capacity of the railway lines and nodes [Liu 2011]. From the mathematical point of view, the stochastic expression is utilized with unknown quantities that are mutually dependent on each other [Kontaxi & Ricci 2010]. On the other hand, the deterministic expression is used with unknown quantities that are mutually independent of each other [Kontaxi & Ricci 2010].

With the analytical method, the infrastructure is studied and calculated with the track lines and railway nodes, which are modeled with suitable queuing systems as single servers or multi-servers. For track lines subdivided into different sections, single severs are used for each direction when applying the analytical method. In addition, in order to measure the capacity of a railway node, it is necessary to analyze the structures of the node in the railway networks as well as the smaller infrastructure elements that make up a railway node.

The railway nodes are the points of the network connections where multiple railway lines are linked. The structure of a node consists of two components [Pachl 2016]: the set of tracks (Gleisgruppen) and the set of conflicting sub routes (Teilfahrstraßenknoten). The set of tracks in a node is modeled as multi-servers of a queuing system. Comparably, because the conflicting sub routes in the node are exclusive of each other, its characteristics can be modeled as single servers in a queuing system. For complex railway nodes, many sets of conflicting sub routes are assembled to be modeled as a multi-resource queue (see [Omahen 1977], [Green 1984] and [Nießen 2008]). Through mathematical analysis, an expected value of the waiting time and the line exploitation rate can be ascertained deterministically.

Furthermore, the operating program is described mathematically with the analytical method. When using the queuing system, the trains are treated as the customers.. The information from the operating program can be described as a random variable, which is the time interval between the arrival times of two trains, using a suitable distribution (or its variance in the simplest case). The occupation times of a train refer to the service time in the server of a queuing system, which is regarded as a random variable with a proper probability distribution (or its variance in the simplest case). Therefore, the operating program can be represented with a suitable mathematical model.

[International Union of Railways (UIC) 2004] describes the compression method to determine the maximum capacity with the time-distance diagram. Through compressing the blocking time stairways and keeping the minimum line headway and the structure of the trains, the consumed capacity of an infrastructure can be derived. However, this compression method as described in [International Union of Railways (UIC) 2004] can only be used under very large restrictions [Lindner 2011]. The line capacity utilized by an operating program can be simply visualized by compressing the blocking time stairways as close as possible together without any buffer time and with keeping the sequence of trains unchanged [Pachl 2016]. According to [Wendler 2002], it is comprehensible that the evaluation of the railway lines capacity research can be carried out using the compression method.

With its mathematical formulas and algebraic expressions, the analytical method is mainly utilized to identify the preliminary resolutions and reference values for capacity research [Rossetti 2009]. It is helpful to use the analytical method in relatively simple situations, which may require more effort and time to simulate than to solve the problem [Pachl 2014]. It is useful to apply the analytical method for relatively simple, homogeneous track arrangements and separated railway infrastructure sections to evaluate the performance of such restricted areas [Rossetti 2009]. The analytical method is designed to provide efficient results based on scant aggregated input data for evaluating the railway capacity on a homogeneous investigated railway infrastructure with a rough operating program instead of detailed timetables.

2.3.2Simulation Method

Simulation is the modeling of a real object or process based on their considered characteristics. Simulation can be utilized to study and evaluate a real object or process instead of using the original real object or process [Siefer 2014]. The simulation method is realized in a computer-based model of a railway system instead of the real operation [Pachl 2014].

A computer-based model can be set up through determining various attributes of the real system with the help of simulation tools like RailSys2, which is introduced in [Siefer 2014]. In a virtual laboratory for a railway system, the infrastructure and timetables (based on the operating program) can be set and adjusted in various ways according to what is required. The model can be used to study and evaluate the real railway system without much effort of time and cost. It has few limitations and is relatively more cost-efficient than a physical experimental model in a real railway system for research and planning [Siefer 2014].

The simulation method can be used for evaluating the operational performance of a railway system, optimizing and validating the operational performance, and planning timetables in a railway system. The railway system consists of various characteristics of the infrastructure including the tracks, switches, signaling systems, station (node) and lines; as well as the rolling stock [Siefer 2014]. In addition, it also includes the operation of trains with the rules on the infrastructure based on the timetable (operating program), and even the operational perturbations as well as dispatching with an appropriate model in the simulation tool. The model using the simulation method is built much closer to reality and higher accuracy of results is obtained. Capacity research can be carried out through simulating a series of timetables with stepwise-varied traffic flows, the results of the evaluation for an operating program or a specific timetable on a given infrastructure and the quality of operation can be obtained with the simulation method and can be further analyzed for optimization.

According to processing techniques with the help of simulation tools, the simulation method can be classified into two types:

Asynchronous simulation

Synchronous simulation

The difference between these two types is in the modeling of the railway operations of the trains. In asynchronous simulation, the trains are based on their priority and the temporal sequence in succession without changing the blocking time stairways (the position of the train paths can be changed but not their slopes in the time-distance-diagram). All of the trains in asynchronous simulation are modeled in the whole process with the help of simulation tool such as LUKS3.

In comparison, with synchronous simulation all the trains are simulated simultaneously (synchronously) at each temporal point. The processes of railway operation in synchronous simulation are simulated in real temporal sequences with the help of simulation tools such as RailSys or OpenTrack4. The results of synchronous simulation are relatively closer to the reality of the operation process, whereas asynchronous simulation is primarily used for scheduling purposes. Nowadays, current simulation tools increasingly combine synchronous and asynchronous approaches.

In this research project, the key point is to evaluate the capacity of the urban rail-bound transport under the conditions of urban mixed traffic. Therefore, consideration will be given to the influences caused by the individual vehicles and pedestrians in urban mixed traffic zones with the aid of the simulation method; these influences are complex and stochastic. For a railway system, [Potthoff 1972] has already proposed the basic stochastic influences in the operational process of railway, which lead to a deviation between the planned and the actual operations. Stochastic influences are regarded as the disturbances randomly caused during railway operation.

The waiting time function is an important intermediate result of simulation method. It was the basis for developing the theory for capacity research of double-track railway lines, while determining the recommended area of traffic flow. It was first proposed by Hertel and Ludwig [Hertel et al. 1987]. The determination of the recommended area of traffic flow with the waiting time function is described in [Hertel 1992]. The theory of [Hertel 1992] allows the recommended area of traffic flow to be derived directly from the waiting time function. To obtain a reasonable and plausible result, the waiting time function has to be determined with sufficient accuracy. In the simulation method, the simulation model of the investigated area has to be suitably modeled. Moreover, the operating program or a scheduled timetable with train types and train mixture is also needed for the basis for the simulation.

Improvements to approximating the waiting time function with a linear approach was put forward and firstly used in the simulation method of railway operation [Bosse et al. 1995]. In order to carry out capacity research with the simulation method, the software PULEIV5