서브메뉴
검색
Curb-Aware Network Models Leveraging Heterogeneous Data: Learning and Optimization
Curb-Aware Network Models Leveraging Heterogeneous Data: Learning and Optimization
상세정보
- 자료유형
- 학위논문 서양
- 최종처리일시
- 20260202105304
- ISBN
- 9798270220716
- DDC
- 620
- 저자명
- Liu, Jiachao.
- 서명/저자
- Curb-Aware Network Models Leveraging Heterogeneous Data: Learning and Optimization
- 발행사항
- [Sl] : Carnegie Mellon University, 2025
- 발행사항
- Ann Arbor : ProQuest Dissertations & Theses, 2025
- 형태사항
- 216 p
- 주기사항
- Source: Dissertations Abstracts International, Volume: 87-06, Section: A.
- 주기사항
- Advisor: Qian, Sean.
- 학위논문주기
- Thesis (Ph.D.)--Carnegie Mellon University, 2025.
- 초록/해제
- 요약Curb space plays an increasingly indispensable role in urban transportation infrastructure systems, serving as the primary interface where multi-modal transportation modes meet, conflict, and compete. With the rapid expansion of new mobility technologies in recent decades, including transportation network company services, on-demand deliveries, micro-mobility, electric vehicles, growing curb usage demand has intensified competition for these limited public assets, bringing negative systematic externalities. Understanding the utilization, occupancy, pricing, and preferred access to curb spaces and the resulting systematic implications requires a integrated network modeling framework for coupled curb and road infrastructure system. Recent advances in sensing and communication technologies have generated vast amount of high-resolution data from various sources, presenting unprecedented opportunities to address these challenges in managing curb infrastructure and transportation systems. Although multi-source data are available, most remain underutilized and isolated within specific zones or applications. Bridging this gap requires a theoretical framework to integrate multi-source data together, unlocking the full potential of big data and enabling smart decision-making across networks and systems.This dissertation presents a comprehensive theoretical framework to model the coupled curb and road infrastructure systems, capturing dynamic curb activities and the intricate relationships between various curb users and their impacts on overall system performance in both static and dynamic networks. The framework leverages large-scale, heterogeneous data to understand spatio-temporal curb usage behavior and network flow patterns. The following major research questions are answered throughout the dissertation: 1) How to characterize spatio-temporal curb usage behaviors and the externalities associated with increasing curb demand in both static and dynamic network models?2) How to evaluate various curb policies and predict travelers' responses in the curb-ware network models and design curb management strategies for certain users to mitigate negative system-wide impacts? 3) How to effectively leverage large-scale, heterogeneous data to infer spatio-temporal curb usage pattern and network flow dynamics? 4) How to enhance model accuracy, reliability and scalability by incorporating large-scale emerging sensing data and advanced calibration methods?Specifically, a static network model is proposed to model multi-modal curb usage and optimal curbside pricing strategy is derived to regulate specific users for system efficiency. A bi-modal user equilibrium (BMUE) model is developed to model mode choice and curb choice of two competitive curb users: private driving and ride-hailing in a general static transportation network, and a curbside queuing model is employed in the equilibrium model to encapsulate the network traffic effect of extensive curbside stopping. Based on the BMUE, an optimal curb pricing design is introduced to regulate ride-hailing curb stops to minimize the system social cost.Next this dissertation proposes a framework of modeling curb usage of heterogeneous users in general dynamic networks. A curb-aware multi-modal dynamic user equilibrium model (C-MMDUE) is proposed to model curb and route choices for three primary curb users. Refined curb space searching and usage dynamics are integrated into the mesoscopic DNL for estimating the externalities of curb usage. Furthermore, a computation graph-based framework is proposed to learn network demand patterns using heterogeneous data including emerging curb event monitoring data together with traditional traffic count and speed data.The model calibration framework is further extended by incorporating remote sensing data (i.e., high-resolution satellite imagery) into DODE to enhance DODE performance with consistent network-wide on-road and curbside parking information. To improve the gradient analysis in heterogeneous traffic flow, an advanced analytical path-based marginal cost is developed, considering heterogeneous traffic flow characteristics and non-differentiability issue, to enhance multi-class travel time gradient analysis. This advanced heterogeneous PMC is applied in solving general system optimum DTA problems.
- 일반주제명
- Engineering
- 일반주제명
- Transportation
- 키워드
- DTA problems
- 키워드
- Curb space plays
- 키워드
- Traffic count
- 키워드
- Speed data
- 기타저자
- Carnegie Mellon University Civil and Environmental Engineering
- 기본자료저록
- Dissertations Abstracts International. 87-06A.
- 전자적 위치 및 접속
- 로그인 후 원문을 볼 수 있습니다.
MARC
008260126s2025 us c eng d■001000017360107
■00520260202105304
■006m o d
■007cr#unu||||||||
■020 ▼a9798270220716
■035 ▼a(MiAaPQ)AAI32282972
■040 ▼aMiAaPQ▼cMiAaPQ
■0820 ▼a620
■1001 ▼aLiu, Jiachao.
■24510▼aCurb-Aware Network Models Leveraging Heterogeneous Data: Learning and Optimization
■260 ▼a[Sl]▼bCarnegie Mellon University▼c2025
■260 1▼aAnn Arbor▼bProQuest Dissertations & Theses▼c2025
■300 ▼a216 p
■500 ▼aSource: Dissertations Abstracts International, Volume: 87-06, Section: A.
■500 ▼aAdvisor: Qian, Sean.
■5021 ▼aThesis (Ph.D.)--Carnegie Mellon University, 2025.
■520 ▼aCurb space plays an increasingly indispensable role in urban transportation infrastructure systems, serving as the primary interface where multi-modal transportation modes meet, conflict, and compete. With the rapid expansion of new mobility technologies in recent decades, including transportation network company services, on-demand deliveries, micro-mobility, electric vehicles, growing curb usage demand has intensified competition for these limited public assets, bringing negative systematic externalities. Understanding the utilization, occupancy, pricing, and preferred access to curb spaces and the resulting systematic implications requires a integrated network modeling framework for coupled curb and road infrastructure system. Recent advances in sensing and communication technologies have generated vast amount of high-resolution data from various sources, presenting unprecedented opportunities to address these challenges in managing curb infrastructure and transportation systems. Although multi-source data are available, most remain underutilized and isolated within specific zones or applications. Bridging this gap requires a theoretical framework to integrate multi-source data together, unlocking the full potential of big data and enabling smart decision-making across networks and systems.This dissertation presents a comprehensive theoretical framework to model the coupled curb and road infrastructure systems, capturing dynamic curb activities and the intricate relationships between various curb users and their impacts on overall system performance in both static and dynamic networks. The framework leverages large-scale, heterogeneous data to understand spatio-temporal curb usage behavior and network flow patterns. The following major research questions are answered throughout the dissertation: 1) How to characterize spatio-temporal curb usage behaviors and the externalities associated with increasing curb demand in both static and dynamic network models?2) How to evaluate various curb policies and predict travelers' responses in the curb-ware network models and design curb management strategies for certain users to mitigate negative system-wide impacts? 3) How to effectively leverage large-scale, heterogeneous data to infer spatio-temporal curb usage pattern and network flow dynamics? 4) How to enhance model accuracy, reliability and scalability by incorporating large-scale emerging sensing data and advanced calibration methods?Specifically, a static network model is proposed to model multi-modal curb usage and optimal curbside pricing strategy is derived to regulate specific users for system efficiency. A bi-modal user equilibrium (BMUE) model is developed to model mode choice and curb choice of two competitive curb users: private driving and ride-hailing in a general static transportation network, and a curbside queuing model is employed in the equilibrium model to encapsulate the network traffic effect of extensive curbside stopping. Based on the BMUE, an optimal curb pricing design is introduced to regulate ride-hailing curb stops to minimize the system social cost.Next this dissertation proposes a framework of modeling curb usage of heterogeneous users in general dynamic networks. A curb-aware multi-modal dynamic user equilibrium model (C-MMDUE) is proposed to model curb and route choices for three primary curb users. Refined curb space searching and usage dynamics are integrated into the mesoscopic DNL for estimating the externalities of curb usage. Furthermore, a computation graph-based framework is proposed to learn network demand patterns using heterogeneous data including emerging curb event monitoring data together with traditional traffic count and speed data.The model calibration framework is further extended by incorporating remote sensing data (i.e., high-resolution satellite imagery) into DODE to enhance DODE performance with consistent network-wide on-road and curbside parking information. To improve the gradient analysis in heterogeneous traffic flow, an advanced analytical path-based marginal cost is developed, considering heterogeneous traffic flow characteristics and non-differentiability issue, to enhance multi-class travel time gradient analysis. This advanced heterogeneous PMC is applied in solving general system optimum DTA problems.
■590 ▼aSchool code: 0041.
■650 4▼aEngineering
■650 4▼aEnvironmental engineering
■650 4▼aTransportation
■653 ▼aBi-modal user equilibrium
■653 ▼aDTA problems
■653 ▼aCurb space plays
■653 ▼aUrban transportation
■653 ▼aTraffic count
■653 ▼aSpeed data
■690 ▼a0543
■690 ▼a0537
■690 ▼a0775
■690 ▼a0709
■71020▼aCarnegie Mellon University▼bCivil and Environmental Engineering.
■7730 ▼tDissertations Abstracts International▼g87-06A.
■790 ▼a0041
■791 ▼aPh.D.
■792 ▼a2025
■793 ▼aEnglish
■85640▼uhttp://www.riss.kr/pdu/ddodLink.do?id=T17360107▼nKERIS▼z이 자료의 원문은 한국교육학술정보원에서 제공합니다.
