東華大學圖書館 |

Algorithms for Real-Time Optimization of Transport Operations in Urban Networks.

紀錄類型:	書目-電子資源 : Monograph/item
正題名/作者:	Algorithms for Real-Time Optimization of Transport Operations in Urban Networks./
作者:	Li, Li.
出版者:	Ann Arbor : ProQuest Dissertations & Theses, : 2021,
面頁冊數:	186 p.
附註:	Source: Dissertations Abstracts International, Volume: 83-07, Section: A.
Contained By:	Dissertations Abstracts International83-07A.
標題:	Transportation. -
電子資源:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=28720328
ISBN:	9798762111874

Algorithms for Real-Time Optimization of Transport Operations in Urban Networks.
Li, Li.

Algorithms for Real-Time Optimization of Transport Operations in Urban Networks. - Ann Arbor : ProQuest Dissertations & Theses, 2021 - 186 p.

Source: Dissertations Abstracts International, Volume: 83-07, Section: A.

Thesis (Ph.D.)--New York University Tandon School of Engineering, 2021.

This item must not be sold to any third party vendors.

Taking the network perspective is paramount in contemporary traffic management since a problem in one part of a network is sometimes best addressed by a decision at another location. Because of the uncertainty in both demand and supply, offline traffic management tools fail to adapt to real-time changes in the system. Robust online control that can deal with unexpected disturbances has become a critical need in today's urban traffic systems. However, most existing online control tools are either centralized and hence computationally prohibitive as the network gets larger, or decentralized using heuristics that usually do not come with any performance guarantees. There is a significant research gap in the online traffic network management.This dissertation aims to develop online network optimization algorithms that can be implemented in real-time in large networks, and at the same time, come with theorectically provable performance guarantees. New technologies such as connected/ automated vehicles (CAV) that have developed fast in recent years have brought or will bring both oppotunities and chanllenges to traffic network management. Assuming a CAV environment, this dissertation focuses on two traffic management tools operated by public sector and private enterprises, respectively. One is the urban network traffic signal control as a representative of tools managed by the public sector, and the second is shared automated electric vehicle (SAEV) systems as the representative of traffic management tools operated by the private sector. The reason for the distinction is that the goals of traffic management do depend on who operates the system.The first part of this study aims to design a computationally efficient real-time signal timing algorithm for large urban networks with theoretical guarantees of performance. According to data from the US Department of Transportation in 2019, approximately 50% of road congestion is caused by temporary disruptions mainly including incidents, work zones, adverse weather, and special events. These discruptions can hardly be anticipated and they dramatically reduce road capacity and system reliability.Traffic signal control algorithms have increased in sophistication over the past decades, from isolated intersection control to corridor coordination, and then to network optimization. Connected vehicle techonologies allow signal controllers to obtain detailed traffic information at low cost to the system, making real-time signal optimization more practical. Decentralized intersection control techniques have received recent attention in the literature as means to overcome scalability issues associated with network-wide intersection control. Chief among these techniques are backpressure (BP) control algorithms, which were originally developed for large wireless networks. In addition to being light-weight computationally, they come with guarantees of performance at the network level, specifically in terms of network-wide stability. The dynamics in BP control are represented using networks of point queues and this also applies to all of the applications to traffic control. As such, BP in traffic fail to capture the spatial distribution of vehicles along the intersection links and, consequently, spill-back dynamics.This dissertation develops a position weighted backpressure (PWBP) control policy for network traffic by applying continuum modeling principles of traffic dynamics, which can capture the spatial distribution of vehicles along network roads and, hence, spill-back dynamics. PWBP inherits the computational advantages of traditional BP. To prove stability of PWBP, (i) a Lyapunov functional that captures the spatial distribution of vehicles is developed; (ii) the capacity region of the network is formally defined in the context of macroscopic network traffic; and (iii) it is proved, when exogenous arrival rates are within the capacity region, that PWBP control is network-wide stabilizing. Comparisons are conducted against a real-world adaptive control implementation for an isolated intersection. Comparisons are also performed against other BP approaches in addition to optimized fixed timing control at the network level. These experiments demonstrate the superiority of PWBP over the other control policies in terms of capacity region, network-wide delay, congestion propagation speed, recoverability from heavy congestion (outside of the capacity region), and response to incidents.Both BP and PWBP are based on an assumption of perfect knowledge of traffic conditions throughout the network at all times, specifically the queue lengths (more accurately, the traffic volumes). However, it has been well established that accurate queue length information at signalized intersections is only available in fully connected environments. Although connected vehicle technologies are developing quickly, we are still far from a fully connected environment even in cities with the most advanced technological infrastructure. The second part of this study hence aims to test the effectiveness of BP/PWBP controls when incomplete or imperfect knowledge about traffic conditions is available. BP/PWBP control are combined with a speed/density field estimation module suitable for a partially connected environment, and the proposed system is referred to as BP/PWBP with estimated queue lengths (BP/PWBP-EQ). The robustness of BP/PWBP-EQ to varying of connected vehicle penetration levels are tested along with comparisons between BP/PWBP-EQ and the original BP/PWBP (i.e., assuming accurate knowledge of traffic conditions), a real-world adaptive signal controller, and optimized fixed timing control using microscopic traffic simulation with field calibrated data. The results show that with a connected vehicle penetration rate as little as 10%, BP/PWBP-EQ can outperform the adaptive controller and the fixed timing controller in terms of average delay, throughput, and maximum stopped queue lengths under high demand scenarios.The third part of this study aims to design a real-time vehicle dispatching algorithm for SAEV systems that comes with network stability and desired dispatch costs. Car-sharing has emerged as a competitive technology for urban mobility. Combined with the upward trend in vehicle electrification and the promise of automation, it is expected that urban travel will change in fundamental ways in the near future. Indeed, breakthroughs in battery technology and the incentive programs offered by governments worldwide have resulted in a continued increase in the market share of electric vehicles. Automation frees passengers from having to drive and seek parking, it also offers increased flexibility when selecting pick up locations. These trends and incentives naturally suggest that SAEV systems will displace traditional gasoline-powered, human-driven car-sharing systems worldwide.Real-time vehicle dispatching operations in traditional car-sharing systems is an already computationally challenging scheduling problem. Electrification only exacerbates the computational difficulties as charge level constraints come into play. To overcome this complexity, the dissertation employs an online minimum drift plus penalty (MDPP) approach for SAEV systems that (i) does not require a priori knowledge of customer arrival rates to the different parts of the system (i.e. it is practical from a real-world deployment perspective), (ii) ensures the stability of customer waiting times, (iii) ensures that the deviation of dispatch costs from a desirable dispatch cost can be controlled, and (iv) has a computational time-complexity that allows for real-time implementation. Using an agent-based simulator developed for SAEV systems, this study tests the MDPP approach under two scenarios with real-world calibrated demand and charger distributions: 1) a low-demand scenario with long trips, and 2) a high-demand scenario with short trips.

ISBN: 9798762111874Subjects--Topical Terms:

555912
Transportation.
Subjects--Index Terms:

Lyapunov optimization

Algorithms for Real-Time Optimization of Transport Operations in Urban Networks.
LDR:09426nmm a2200373 4500 001 2345038
005 20220531062341.5
008 241004s2021 ||||||||||||||||| ||eng d
020 $a 9798762111874
035 $a (MiAaPQ)AAI28720328
035 $a AAI28720328
040 $a MiAaPQ $c MiAaPQ
100 1 $a Li, Li. $3 1036040
245 1 0 $a Algorithms for Real-Time Optimization of Transport Operations in Urban Networks.
260 1 $a Ann Arbor : $b ProQuest Dissertations & Theses, $c 2021
300 $a 186 p.
500 $a Source: Dissertations Abstracts International, Volume: 83-07, Section: A.
500 $a Advisor: Jabari, Saif-Eddin.
502 $a Thesis (Ph.D.)--New York University Tandon School of Engineering, 2021.
506 $a This item must not be sold to any third party vendors.
520 $a Taking the network perspective is paramount in contemporary traffic management since a problem in one part of a network is sometimes best addressed by a decision at another location. Because of the uncertainty in both demand and supply, offline traffic management tools fail to adapt to real-time changes in the system. Robust online control that can deal with unexpected disturbances has become a critical need in today's urban traffic systems. However, most existing online control tools are either centralized and hence computationally prohibitive as the network gets larger, or decentralized using heuristics that usually do not come with any performance guarantees. There is a significant research gap in the online traffic network management.This dissertation aims to develop online network optimization algorithms that can be implemented in real-time in large networks, and at the same time, come with theorectically provable performance guarantees. New technologies such as connected/ automated vehicles (CAV) that have developed fast in recent years have brought or will bring both oppotunities and chanllenges to traffic network management. Assuming a CAV environment, this dissertation focuses on two traffic management tools operated by public sector and private enterprises, respectively. One is the urban network traffic signal control as a representative of tools managed by the public sector, and the second is shared automated electric vehicle (SAEV) systems as the representative of traffic management tools operated by the private sector. The reason for the distinction is that the goals of traffic management do depend on who operates the system.The first part of this study aims to design a computationally efficient real-time signal timing algorithm for large urban networks with theoretical guarantees of performance. According to data from the US Department of Transportation in 2019, approximately 50% of road congestion is caused by temporary disruptions mainly including incidents, work zones, adverse weather, and special events. These discruptions can hardly be anticipated and they dramatically reduce road capacity and system reliability.Traffic signal control algorithms have increased in sophistication over the past decades, from isolated intersection control to corridor coordination, and then to network optimization. Connected vehicle techonologies allow signal controllers to obtain detailed traffic information at low cost to the system, making real-time signal optimization more practical. Decentralized intersection control techniques have received recent attention in the literature as means to overcome scalability issues associated with network-wide intersection control. Chief among these techniques are backpressure (BP) control algorithms, which were originally developed for large wireless networks. In addition to being light-weight computationally, they come with guarantees of performance at the network level, specifically in terms of network-wide stability. The dynamics in BP control are represented using networks of point queues and this also applies to all of the applications to traffic control. As such, BP in traffic fail to capture the spatial distribution of vehicles along the intersection links and, consequently, spill-back dynamics.This dissertation develops a position weighted backpressure (PWBP) control policy for network traffic by applying continuum modeling principles of traffic dynamics, which can capture the spatial distribution of vehicles along network roads and, hence, spill-back dynamics. PWBP inherits the computational advantages of traditional BP. To prove stability of PWBP, (i) a Lyapunov functional that captures the spatial distribution of vehicles is developed; (ii) the capacity region of the network is formally defined in the context of macroscopic network traffic; and (iii) it is proved, when exogenous arrival rates are within the capacity region, that PWBP control is network-wide stabilizing. Comparisons are conducted against a real-world adaptive control implementation for an isolated intersection. Comparisons are also performed against other BP approaches in addition to optimized fixed timing control at the network level. These experiments demonstrate the superiority of PWBP over the other control policies in terms of capacity region, network-wide delay, congestion propagation speed, recoverability from heavy congestion (outside of the capacity region), and response to incidents.Both BP and PWBP are based on an assumption of perfect knowledge of traffic conditions throughout the network at all times, specifically the queue lengths (more accurately, the traffic volumes). However, it has been well established that accurate queue length information at signalized intersections is only available in fully connected environments. Although connected vehicle technologies are developing quickly, we are still far from a fully connected environment even in cities with the most advanced technological infrastructure. The second part of this study hence aims to test the effectiveness of BP/PWBP controls when incomplete or imperfect knowledge about traffic conditions is available. BP/PWBP control are combined with a speed/density field estimation module suitable for a partially connected environment, and the proposed system is referred to as BP/PWBP with estimated queue lengths (BP/PWBP-EQ). The robustness of BP/PWBP-EQ to varying of connected vehicle penetration levels are tested along with comparisons between BP/PWBP-EQ and the original BP/PWBP (i.e., assuming accurate knowledge of traffic conditions), a real-world adaptive signal controller, and optimized fixed timing control using microscopic traffic simulation with field calibrated data. The results show that with a connected vehicle penetration rate as little as 10%, BP/PWBP-EQ can outperform the adaptive controller and the fixed timing controller in terms of average delay, throughput, and maximum stopped queue lengths under high demand scenarios.The third part of this study aims to design a real-time vehicle dispatching algorithm for SAEV systems that comes with network stability and desired dispatch costs. Car-sharing has emerged as a competitive technology for urban mobility. Combined with the upward trend in vehicle electrification and the promise of automation, it is expected that urban travel will change in fundamental ways in the near future. Indeed, breakthroughs in battery technology and the incentive programs offered by governments worldwide have resulted in a continued increase in the market share of electric vehicles. Automation frees passengers from having to drive and seek parking, it also offers increased flexibility when selecting pick up locations. These trends and incentives naturally suggest that SAEV systems will displace traditional gasoline-powered, human-driven car-sharing systems worldwide.Real-time vehicle dispatching operations in traditional car-sharing systems is an already computationally challenging scheduling problem. Electrification only exacerbates the computational difficulties as charge level constraints come into play. To overcome this complexity, the dissertation employs an online minimum drift plus penalty (MDPP) approach for SAEV systems that (i) does not require a priori knowledge of customer arrival rates to the different parts of the system (i.e. it is practical from a real-world deployment perspective), (ii) ensures the stability of customer waiting times, (iii) ensures that the deviation of dispatch costs from a desirable dispatch cost can be controlled, and (iv) has a computational time-complexity that allows for real-time implementation. Using an agent-based simulator developed for SAEV systems, this study tests the MDPP approach under two scenarios with real-world calibrated demand and charger distributions: 1) a low-demand scenario with long trips, and 2) a high-demand scenario with short trips.
520 $a The comparisons with other algorithms under both scenarios show that the proposed online MDPP outperforms all other algorithms in terms of both reduced customer waiting times and vehicle dispatching costs.
590 $a School code: 1988.
650 4 $a Transportation. $3 555912
653 $a Lyapunov optimization
653 $a Network optimization
653 $a Netwrok traffic management
653 $a Position-weighted backpressure
653 $a Shared automated electric vehicle
653 $a Shared mobility
690 $a 0709
710 2 $a New York University Tandon School of Engineering. $b Civil Engineering. $3 3350166
773 0 $t Dissertations Abstracts International $g 83-07A.
790 $a 1988
791 $a Ph.D.
792 $a 2021
793 $a English
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=28720328