- Ms Word Format
- 70 Pages
- ₦3000
- 1-5 Chapters
TRAFFIC SIGNAL CONTROL IN A CONNECTED AND AUTONOMOUS VEHICLE ENVIRONMENT CONSIDERING PEDESTRIANS
TRAFFIC SIGNAL CONTROL IN A CONNECTED AND AUTONOMOUS VEHICLE ENVIRONMENT CONSIDERING PEDESTRIANSABSTRACT Traffic signals help to maintain order in urban traffic networks and reduce vehicle conflicts by dynamically assigning right-of-way to different vehicle movements. However, by temporarily stopping vehicle movements at regular intervals, traffic signals are a major source of urban congestion and cause increased vehicle delay, fuel consumption, and environmental pollution. Connected and Autonomous Vehicle technology may be utilized to optimize traffic operations at signalized intersections, since connected vehicles have the ability to communicate with the surrounding infrastructure and autonomous vehicles can follow the instructions from the signal or a central control system. Connected vehicle information received by a signal controller can be used to help adjust signal timings to tailor to the specific dynamic vehicle demand. Information about the signal timing plan can then be communicated back to the vehicles so that they can adjust their speeds/trajectories to further improve traffic operations. Based on a thorough literature review of existing studies in the area of signal control utilizing information from connected and autonomous vehicles, three research gaps are found: 1) application are limited to unrealistic intersection configurations; 2) methods are limited to a single mode; or, 3) methods only optimize the average value of measure of effectiveness while ignoring the distribution among vehicles. As a part of this dissertation, several methods will be proposed to increase computational efficiency of an existing CAV-based joint signal timing and vehicle trajectory optimization algorithm so that it can be applied to more realistic intersection settings without adding computational burden. Doing so requires the creation of new methods to accommodate features like multiple lanes on each approach, more than two approaches and turning maneuvers. Methods to incorporate human-driven cooperative vehicles and pedestrians are also proposed and tested. A more equitable traffic signal control method is also designed. TABLE OF CONTENTS LIST OF FIGURES ……………………………………………………………………………………………… vi LIST OF TABLES……………………………………………………………………………………………….. ix LIST OF EQUATIONS ………………………………………………………………………………………… x Chapter 1 Introduction …………………………………………………………………………………………. 1 Chapter 2 Signal Timing Optimization with Connected Vehicle Technology: Platooning to Improve Computational Efficiency ……………………………………………………………….. 5 Introduction …………………………………………………………………………………………………. 6 Literature review …………………………………………………………………………………………… 7 Signal control algorithm …………………………………………………………………………………. 10 Step 1: Platoon identification ……………………………………………………………………. 14 Step 2: Optimal platoon departure sequence selection …………………………………… 15 Step 3: Longitudinal trajectory guidance …………………………………………………….. 16 Intersections with two single-lane approaches…………………………………………………….. 17 Simulation framework …………………………………………………………………………….. 18 Sensitivity to critical headway and spacing threshold values ………………………….. 19 Intersections with four multi-lane approaches and conflicting left turns ………………….. 23 Extension to accommodate more complex intersection configurations ……………… 23 Simulations……………………………………………………………………………………………. 25 Conclusion …………………………………………………………………………………………………… 28 Chapter 3 Joint Optimization of Signal Phasing and Timing and Vehicle Speed Guidance in a Connected and Autonomous Vehicle Environment………………………………………… 31 Introduction …………………………………………………………………………………………………. 32 Methodology ………………………………………………………………………………………………… 35 Framework of the control algorithm …………………………………………………………… 35 Intersection configuration and vehicle types ………………………………………………… 36 Signal optimization method ……………………………………………………………………… 37 Speed guidance design …………………………………………………………………………….. 40 Simulation tests …………………………………………………………………………………………….. 43 Sensitivity of control algorithm to driver characteristics ………………………………………. 51 Sensitivity of control frequency ……………………………………………………………………….. 56 Conclusion and future work…………………………………………………………………………….. 58 Chapter 4 An Equitable Traffic Signal Control Scheme at Signalized Intersections Using Connected Vehicle Technology ……………………………………………………………………….. 61 Introduction …………………………………………………………………………………………………. 62 Literature review …………………………………………………………………………………………… 63 Methodology ………………………………………………………………………………………………… 67Scenario considered ………………………………………………………………………………… 67Proposed signal control algorithm ……………………………………………………………… 68 Simulation and results ……………………………………………………………………………………. 74 Simulation platform and scenario ………………………………………………………………. 74 Vehicle delay distribution without maximum delay threshold …………………………. 75 Vehicle delay distribution with maximum delay threshold……………………………… 78 Tradeoff between delay equity and efficiency ……………………………………………… 82 Sensitivity of the proposed method to demand …………………………………………….. 85 Analysis of imperfect market penetration of connected vehicles ……………………………. 90Conclusions …………………………………………………………………………………………………. 94 Chapter 5 A Heuristic Method to Optimize Generic Signal Phasing and Timing Plans at Signalized Intersections Using Connected Vehicle Technology …………………………….. 97 Introduction …………………………………………………………………………………………………. 98 Intersection signal control in a connected vehicle environment ……………………….. 98 Signal timing optimization using genetic algorithm methods ………………………….. 100 Summary and outline of paper ………………………………………………………………….. 101 Methodology ………………………………………………………………………………………………… 103 Intersection layout and vehicle types ………………………………………………………….. 103 Signal control algorithm ………………………………………………………………………….. 104 Solution methods ………………………………………………………………………………………….. 107 Enumeration approaches ………………………………………………………………………….. 107 Genetic algorithm approaches …………………………………………………………………… 110 Simulation and results ……………………………………………………………………………………. 116 Comparison between four-phase plan and eight-phase plan using complete enumeration ……………………………………………………………………………………. 116Performance of all heuristic control methods ……………………………………………….. 119 Sensitivity of analysis methods to intersection properties ………………………………. 122 Conclusion …………………………………………………………………………………………………… 125 Chapter 6 Traffic Signal Control Optimization in a Connected Vehicle Environment Considering Pedestrians …………………………………………………………………………………. 127 Introduction …………………………………………………………………………………………………. 128 Methodology ………………………………………………………………………………………………… 131 Intersection configuration and vehicle and pedestrian types ……………………………. 131 Signal control algorithm ………………………………………………………………………….. 134 Simulation tests …………………………………………………………………………………………….. 140 Person-based delay under different weight ratios between vehicle and pedestrian delay ………………………………………………………………………………. 141 Sensitivity of the proposed method to demand pattern …………………………………… 143 Analysis of imperfect arrival information of pedestrians and vehicles …………………….. 146 Imperfect pedestrian information ………………………………………………………………. 147 Imperfect vehicle information …………………………………………………………………… 150 Concluding remarks ………………………………………………………………………………………. 151 Chapter 7 Conclusion…………………………………………………………………………………………… 154 References ………………………………………………………………………………………………………….. 156
Chapter 1
Introduction Urban traffic congestion is a large contributor to travel delays, vehicle emissions, and wasted fuel and the problem has become increasingly more serious over the past few decades. For example, historical data of traffic demands and various types of congestion costs from the year 1982 to 2012 in Pittsburgh shown in Table 1-1 indicate that traffic demand, fuel consumption, traffic delay, and congestion cost generally all increase with time (Texas A&M Transportation Institute, 2014). Table 1-1: Traffic demands and costs in Pittsburgh from 1982 to 2012.
Year | Auto commuters [thousands] | Annual excess fuel consumed [gal/auto commuter] | Annual delay [hr/auto commuter] | Annual congestion cost [$/auto commuter] |
1982 | 587 | 5 | 11 | 477 |
1987 | 608 | 8 | 19 | 704 |
1992 | 648 | 12 | 28 | 892 |
1997 | 737 | 15 | 31 | 944 |
2002 | 826 | 18 | 34 | 1,011 |
2007 | 844 | 20 | 37 | 957 |
2012 | 852 | 20 | 37 | 882 |
Signalized intersections temporarily halt vehicle movements at regular intervals, and thus serve as the primary bottlenecks in urban environments that contribute to the negative outcomes previously described (Guler et al., 2014). Based on a recent report from Federal Highway Administration, Traffic Congestion and Reliability: Linking Solutions and Problems, poor signal timing is responsible for 5% of total traffic congestion (Cambridge Systematics, 2004), see Figure 1-1. Although the percentage seems small, poor signal timing constantly contributes to urban congestion, whereas other sources only occur during specific events, such as traffic incidents and work zones. In addition, field tests have shown that stop-and-go driving (that can be caused by poor signal timing) causes 14% more exhaust gas emissions compared to vehicles that drive at a constant speed (Xia et al., 2012). Figure 1-1: Sources of congestion. An effective signal control strategy enables safe vehicle movement while minimizing negative outcomes, such as delays to individual users or number of times vehicles have to stop while proceeding through the intersection. A variety of signal control algorithms have been proposed and implemented to help provide safe vehicle movement while minimizing these other negative outcomes. Existing traffic signal control methods range from pre-timed phasing plans that are optimized a priori and do not vary with prevailing traffic conditions to actuated or adaptive signal control algorithms that adjust signal timing in response to traffic measurements obtained from sensors at the intersection in real-time. These approaches are generally based on macroscopic or aggregate traffic data such as vehicle flows or average occupancies over some time period. A summary of contemporary signal control strategies can be found in Papageorgiou et al. (2003). However, these methods based on macroscopic traffic information cannot account for individual vehicle information, thus are unable to provide dynamic and accurate signal adjustments based on real-time traffic condition. Recent advances in vehicular, communications, and mobile computing technologies offer the potential for even more flexible traffic signal control paradigms. New communication technologies allow vehicles to collect and share more detailed information about existing traffic patterns. The emergence of the connected vehicle (CV) technology can provide more detailed information about vehicle arrival patterns at an intersection that can be used to design more effective and dynamic signal control schemes. The introduction of autonomous vehicles (AVs) will also allow vehicles and signals to cooperate as a means to improve signalized intersection operations. One recent line of research has proposed using real-time information from connected and automated vehicles (CAVs) to minimize total vehicle delay at signalized intersections by optimizing individual vehicle departure sequences (Guler et al., 2014; Yang et al., 2015; Yang et al., 2018) while also minimizing total vehicle stopping maneuvers by modifying the longitudinal trajectories of AVs (Yang et al., 2016). However, these algorithms were developed only for single, isolated intersections with two single-lane approaches that could be controlled with simple two-phase signals and did not have conflicting left turns. Extension to more realistic intersections (e.g., those with four approaches, multiple lanes on each approach, or conflicting left-turning maneuvers) with more vehicles and phasing options was not possible due to the computational demands imposed. Moreover, only one traffic mode is considered, which limits its application in real world scenarios; and only the average delay is considered in optimization without consideration of the distribution of vehicle delays in the fleet. In light of this, this dissertation develops methods to extend this algorithm to more equitable applications and more realistic situations, including: intersection geometries with more than two approaches, multi-lane approaches, turning traffic and multiple modes. The remainder of this dissertation includes the five journal papers published or submitted through the research of this topic, with a conclusion remark presented in the last chapter. The five papers focus on: 1) improving computation efficiency by grouping vehicles into naturally occurring platoons; 2) designing vehicle speed guidance for connected and autonomous vehicles for reducing number of stops; 3) proposing a more equitable signal control scheme besides minimizing average vehicle delay; 4) developing heuristic method to extend the signal control algorithm to more complicated intersection configurations; 5) incorporating pedestrian delay and presenting a person-based signal control algorithm; respectively. TRAFFIC SIGNAL CONTROL IN A CONNECTED AND AUTONOMOUS VEHICLE ENVIRONMENT CONSIDERING PEDESTRIANS