_ap_ufes{"success":true,"siteUrl":"projectchampionz.com.ng","urls":{"Home":"https://projectchampionz.com.ng","Category":"https://projectchampionz.com.ng/projecttopics/about-project-championz/","Archive":"https://projectchampionz.com.ng/2022/05/","Post":"https://projectchampionz.com.ng/2022/05/17/human-factors-as-determinants-of-marine-accidents-in-maritime-companies-in-nigeria/","Page":"https://projectchampionz.com.ng/pathology-project-topics-and-materials/","Attachment":"https://projectchampionz.com.ng/2022/04/27/does-making-money-online-require-software-tools/ppp2/","Nav_menu_item":"https://projectchampionz.com.ng/2021/10/10/93928/","Custom_css":"https://projectchampionz.com.ng/2019/10/18/proc-2/","Oembed_cache":"https://projectchampionz.com.ng/2020/08/09/010e02649ccd61f4b1233b15599ed153/","Wp_global_styles":"https://projectchampionz.com.ng/2022/04/06/wp-global-styles-proc/","Amp_validated_url":"https://projectchampionz.com.ng/amp_validated_url/15a6906a18bf1e2723b7d6bb7c1a58c0/","Buttons-x":"https://projectchampionz.com.ng/buttons-x/flat-info-2/","Acf-field-group":"https://projectchampionz.com.ng/?post_type=acf-field-group&p=56510","Acf-field":"https://projectchampionz.com.ng/?post_type=acf-field&p=56514","Wpcf7_contact_form":"https://projectchampionz.com.ng/?post_type=wpcf7_contact_form&p=2077","Rpt_pricing_table":"https://projectchampionz.com.ng/rpt_pricing_table/price-list-1/","Wp_automatic":"https://projectchampionz.com.ng/wp_automatic/iproject/","Is_search_form":"https://projectchampionz.com.ng/?post_type=is_search_form&p=70552"}}_ap_ufee USING MOBILE PROBE DATA TO ESTIMATE TRAFFIC STATES AND THE EFFECTS OF PERIMETER CONTROL ON URBAN NETWORKS - Project Topics for Student

USING MOBILE PROBE DATA TO ESTIMATE TRAFFIC STATES AND THE EFFECTS OF PERIMETER CONTROL ON URBAN NETWORKS

  • : Ms Word Format
  • : 70 Pages
  • : ₦3000
  • : 1-5 Chapters
  •  
  • Click to DOWNLOAD Materials

USING MOBILE PROBE DATA TO ESTIMATE TRAFFIC STATES AND THE EFFECTS OF PERIMETER CONTROL ON URBAN NETWORKS

ABSTRACT

 

Aggregate models of traffic on urban networks have been studied for decades extending back to the 1960s and various theories have been developed to describe networks as a function of measurable vehicular properties. Earlier models found that a monotonically decreasing relationship existed between the average speed and flow in the network. However, a monotonically decreasing relationship is only practical under freeflow or light conditions; therefore, these earlier models were only able to describe the uncongested regime. Recent work has shown that the average flow and density of a network are related and can dynamically describe both uncongested and congested conditions in an urban network. This relationship, known as the Macroscopic

Fundamental Diagram (MFD), can be used to describe large-scale network dynamics and develop network-wide control strategies to improve efficiency.

This work examined the accuracy of combining data from fixed sensors and mobile probe vehicles to estimate traffic states in real-time. Formulae were developed to estimate the average flow, density, speed, accumulation, and exit flow, along with the uncertainty of these estimates. The methodology was tested using data obtained from several micro-simulated networks to replicate realistic urban transportation networks. Tests using the micro-simulated models showed that estimations were within 10% of the true value for average flow, density, accumulation, and exit flow when 20% of the circulating vehicles served as probes and average speed estimations were shown to be even more accurate. In addition, the mobile probe estimation methodology was used to inform a network-wide perimeter metering strategy on an idealized grid network. These

 

results were compared to metering based on actual traffic states, and the results illustrated that metering can significantly reduce total travel times compared to the non-metering case if 25% of the vehicles in the network serve as probe vehicles. In general, this work shows that the estimation method may be used to inform network-wide metering strategies, which improve network stability and reduce congestion without additional

costly infrastructure improvements.

 

TABLE OF CONTENTS

List of Figures ……………………………………………………………………………………………….. vii

List of Tables ………………………………………………………………………………………………… viii

Acknowledgements ………………………………………………………………………………………… ix

Chapter 1 Introduction ……………………………………………………………………………………. 1

Chapter 2 Literature Review ……………………………………………………………………………. 5

2.1.Background on macroscopic models of urban networks …………………………. 5

2.1.1. Early theories of aggregate level traffic relationships at the city-

scale ………………………………………………………………………………………….. 5

2.1.2. The existence of the macroscopic fundamental diagram on urban

networks ……………………………………………………………………………………. 7

2.1.3. Network-wide control strategies and the macroscopic fundamental

diagram ……………………………………………………………………………………… 12

Chapter 3 Estimation Methodology ………………………………………………………………….. 17

3.1.Metrics of interest ……………………………………………………………………………… 17

3.2.Estimating the metrics of interest using probe data ………………………………… 18

3.3.Estimating the probe penetration rate …………………………………………………… 20

3.4.Accuracy of the estimates …………………………………………………………………… 21

3.4.1. Accuracy of the probe penetration rate ……………………………………….. 21

3.4.2. Accuracy of the metrics of interest ……………………………………………… 223.4.3. Applications using fractional error ……………………………………………… 24

Chapter 4 Tests Using Micro-Simulation ………………………………………………………….. 26

4.1.Idealized network ……………………………………………………………………………… 26

4.2.Orlando network ……………………………………………………………………………….. 28

4.3.Accuracy of estimations …………………………………………………………………….. 31

4.3.1. Accuracy of probe penetration rate …………………………………………….. 31

4.3.2. Comparing probe vehicle data to all vehicle data …………………………. 33

4.3.3. Estimating metrics of interest …………………………………………………….. 344.3.4. Accuracy of variance estimations ……………………………………………….. 384.3.5. Estimating the macroscopic fundamental diagram ………………………… 42

4.3.6. Quantifying the accuracy of MFD estimations …………………………….. 47

4.3.7. Applying estimation methods using fractional error ……………………… 50

4.4.General remarks ………………………………………………………………………………… 51Chapter 5 Network-wide Perimeter Control Strategy ………………………………………….. 53

5.1.Perimeter control strategy …………………………………………………………………… 53

5.2.Traffic state estimations …………………………………………………………………….. 55

5.2.1. Link method …………………………………………………………………………….. 56

5.2.2. Mobile probe method ……………………………………………………………….. 57

5.3.Simulated network …………………………………………………………………………….. 58

5.4.Results …………………………………………………………………………………………….. 60

5.4.1. Non-metered scenarios for link and probe methods ………………………. 62

5.4.2. Metered scenarios for link and probe methods ……………………………… 64

Chapter 6 Conclusions ……………………………………………………………………………………. 69

References …………………………………………………………………………………………………….. 73

Chapter 1

 

Introduction

Traffic congestion is a severe problem in urban transportation networks across the world due to its negative effects on safety, reliability, and the environment. Congestion in 2012 caused urban Americans to travel 5.5 billion hours more, purchase an extra 2.9 billion gallons of fuel for a total cost of $121 billion, and release 56 billion pounds of additional carbon dioxide greenhouse gas into the atmosphere (Schrank et al., 2012). Transportation agencies have attempted to mitigate this problem for decades by constructing new infrastructure to provide more capacity to meet increasing demands. However, expanding existing or constructing new infrastructure in urban areas is infeasible due to limited space and high costs. It has also been shown that road improvements that attempt to mitigate congestion can actually attract traffic from other routes and encourage longer and more frequent trips—a problem commonly known as generated traffic or induced travel (Litman, 2001). This has led to pushes in alternative methods to reduce congestion by utilizing existing infrastructure.

A common method is the use of Intelligent Transportation Systems (ITS) to gain useful information from drivers’ vehicles to monitor traffic conditions. If conditions are accurately measured in real-time, data can be transmitted via communication systems, and ideal control strategies may be implemented. Well-established technologies have existed for several decades that collect data such as vehicle counts, occupancies, travel times, delays, and speeds that are processed and communicated to transportation users via dynamic message signs, traffic signal systems, and road weather information systems. The most common methods are embedded sensors in the infrastructure such as inductive loop detectors or over-roadway sensors such as video image processors. Although these technologies are common practice and are extremely useful to detect vehicles at intersections or to determine freeway travel speeds, they are infrastructure intensive, costly to construct and maintain, and not very effective when vehicles are stopped. The lattermost of these problems tends to arise in urban networks where loop detectors are placed near intersections. Often in these situations, queues spill back over the detector, which causes inaccurate measurements (Buisson and Ladier, 2009; Geroliminis and Sun, 2011b; Courbon and Leclercq, 2011). In addition, these technologies are typically managed on smaller scales—i.e., along a specific coordinated corridor or section of freeway—and are not used to describe network-wide macroscopic properties.

Recent advancements in technologies such as Global Positioning System (GPS) devices provide a new type of sensor that eases the collection of traffic data. These technologies do not require additional infrastructure, greatly reduce the problems previously discussed, and are becoming increasingly common in fleet vehicles and private automobiles. The GPS devices collect vehicle information such as time, speed, and acceleration that can be aggregated over time to calculate network properties for realtime traffic prediction. In addition, the Federal Highway Administration (FHWA) is leading research on the Connected Vehicle program (U.S. Department of Transportation, 2013) that utilizes advanced wireless communications and smart infrastructure that can be directly used to collect vehicle data.

Most research in the field of mobile probe data focuses primarily on estimating traffic properties on arterial roadways (Srinivasan and Jovanis, 1996) or speeds on freeways (Hofleitner et al., 2012) and do not consider urban-scale networks. However, recently Daganzo (2007) proposed that a unique and reproducible relationship exists between the average flow and density of vehicles in a small urban neighborhood or network. This relationship, now commonly known as the Macroscopic or Network Fundamental Diagram (MFD or NFD), provides a feasible option to utilize GPS data to predict urban network conditions in real-time. The MFD can be used to describe the dynamics of urban traffic networks that unveil interesting phenomena and control policies. For example, it was demonstrated using this macroscopic model that urban networks have an innate tendency towards gridlock when congested, and an optimal control scheme was developed to improve network efficiency by metering vehicle entry into the network (Daganzo, 2007; Geroliminis and Daganzo, 2008).

The MFD must be estimated before it can be applied to describe the behavior of a network and implement control. Estimation of the MFD is a highly data-intensive process, and only few studies have been able to derive MFDs from empirical data (Geroliminis and Daganzo, 2008; Buisson and Ladier, 2009). These successful studies relied heavily on loop detector data, whose placement have been shown to significantly affect macroscopic relationships (Buisson and Ladier, 2009; Geroliminis and Sun, 2011b; Courbon and Leclereq, 2011). Although Keyvan-Ekbatani et al. (2012) showed that limited detector information is sufficient to implement gating control on a specific network, the placement of detectors may provide inaccurate MFDs. Since network control strategies such as pricing are highly sensitive, more accurate MFDs are required.

Gayah and Dixit (2013) developed a method to estimate average network densities in real-time by combining probe vehicle data with a known MFD. However, it requires that the MFD is known, it cannot predict traffic states in light flow conditions, and it cannot be used if the MFD exhibits any hysteresis. As an alternative approach, this current work provides a general method to accurately estimate average network flow, density, speed, accumulation, and exit flow under both uncongested free-flow states and congested states without prior knowledge of the MFD. This method allows for GPS data to be used to derive the MFD and monitor traffic conditions in real-time. With accurate state estimations, the information may be used to inform vehicle re-routing schemes and network-wide traffic control strategies to improve network stability and performance.

The objectives of this work are to first determine a methodology that combines mobile probe vehicle data with macroscopic traffic models for network-wide traffic state estimations. Next, the uncertainties of the traffic state estimations are to be quantified. To confirm the accuracy of the estimations and uncertainties, an idealized grid network and a more realistic simulated model of downtown Orlando, FL are used. Additionally, macroscopic traffic control strategies that utilize mobile probe data to improve overall network efficiency are developed. These strategies are assessed using the idealized micro-simulation model by comparing the effectiveness of the mobile probe data estimations with data from all vehicles in the network. Lastly, the practicality of the traffic state estimations is evaluated.

USING MOBILE PROBE DATA TO ESTIMATE TRAFFIC STATES AND THE EFFECTS OF PERIMETER CONTROL ON URBAN NETWORKS

Sharing is caring!

Leave a Reply