A5042528213- Transportation Planning and Optimization
- Traffic control and management
- Traffic Prediction and Management Techniques
- Transportation and Mobility Innovations
- Urban Transport and Accessibility
- Traffic and Road Safety
- Evacuation and Crowd Dynamics
- Neural Networks and Applications
- Evolutionary Algorithms and Applications
- Image Retrieval and Classification Techniques
- Reinforcement Learning in Robotics
- Advanced Graph Neural Networks
- Recommender Systems and Techniques
- Time Series Analysis and Forecasting
- Complex Systems and Time Series Analysis
- Economic and Environmental Valuation
- Artificial Intelligence in Games
Hangzhou Dianzi University
2024-2025
University of California, Davis
2000-2015
Tongji University
2004-2015
University of California System
2007-2009
As a typical self-driven many-particle system far from equilibrium, traffic flow exhibits diverse fascinating non-equilibrium phenomena, most of which are closely related to stability and specifically the growth/dissipation pattern disturbances. However, theories have been controversial due lack precise data. We studied new perspective by carrying out large-scale car-following experiment on an open road section, overcomes intrinsic deficiency empirical observations. The has shown clearly...
This paper studies dynamic network models in which congestion takes the form of queuing behind bottlenecks. The three consideration, point queue (P-Q) model, spatial (S-Q) model and cell transmission all deal with queues caused by temporal bottlenecks, yet differ each other describing S-Q characteristics. Our work focuses on exploring impacts distribution result loading (DNL). For this purpose, simple node (merge diverge) are introduced to describe flow interactions across links. We present...
This paper presents a recursive traffic flow prediction algorithm using artificial neural networks. The system model is specified based on the understanding of how disturbances in are propagated, and order determined by correlation analysis. parameters model, other hand, can be obtained through nonlinear optimization. Preliminary studies show that this approach yield reasonably accurate results.
The cell-based system optimal dynamic traffic assignment (SO-DTA) model has recently been applied to study emergency evacuation by a handful of authors. It is recognized that an solution this may contain phenomenon called holding, which discharges flow at lower rate than what can be achieved under the given conditions. Mathematically, caused relaxation propagation constraints. In paper, pattern contains no holding always proved exist in context planning. An without much easier and less...
Congestion pricing, one of the effective tools to reduce congestion in a transportation network, may cause inequity among commuters if differences their value time (VOT) are not properly taken into account. In this paper, we develop bimodal competition model within context nonidentical and departure choice study toll design, mode share, benefit distribution problems. We first show that for single bottleneck without schedule-late delay, pass increasing order VOT under an optimal dynamic toll,...
Traffic flow complexity comes from the car-following and lane-changing behavior. Based on empirical data for individual vehicle speeds time headways measured a single-lane highway section, we have studied traffic properties induced by pure We found that spontaneous sudden drop in velocity could happen platoon of vehicles when leading is quite high (\ensuremath{\sim}70 km/h). In contrast, slows down, such has not been observed. Our finding indicates breakdown road might be phase transition...
Freeway lane drops are typical sites where long queues and stop–start traffic waves can be observed. A theory was recently developed to explain the formation of oscillations upstream a freeway drop-in. Using that theory, this paper carries out numerical investigation explore how parameters characterizing driving behavior at geometry lane-drop site may affect these oscillatory patterns. The effects five factors–-distance between merging sign taper, travel demand, remaining capacity, overall...
To obtain more credible estimates of time-dependent travel demand, various data sources should be exploited jointly to improve the observability origin–destination (O-D) trip tables. A comprehensive case study uses a real freeway network reveal how different coverage affects quality estimated O-D The dynamic estimation problem is formulated as variational inequality (VI) that provides flexible framework incorporate and encapsulate realistic traffic flow dynamics. Traffic surveillance...
Traffic oscillations, an unpleasant form of traffic congestion, have attracted the attention researchers for years. When viewed in so-called flow–density or speed–density phase plane, oscillatory states often present themselves hysteresis loops. In this paper, impacts driver relaxation and anticipation on flow are examined, their link to is sought. Through analysis trajectory data from NGSIM a theoretical car-following models, it revealed that generated by imbalance anticipation. Changing...
This paper provides a hierarchical framework for studying the impact of traveler information on network reserve capacity. The comprises two-level mathematical program—the upper-level program maximizes capacity multiplier subject to link constraint, and lower-level generates user equilibrium flow patterns under influence information. is solved by genetic algorithm-based solution method. Numerical results indicate that road does not increase monotonically with level. implications this finding,...
We have carried out car-following experiments with a 25-car-platoon on an open road section to study the relation between car's speed and its spacing under various traffic conditions, in hope resolve controversy surrounding this fundamental of vehicular traffic. In paper we extend our previous analysis these experiments, report new experimental findings. particular, reveal that platoon length (hence average within platoon) might be significantly different even if velocity is essentially...
This paper studies three types of algorithms for computing all-to-one time-dependent shortest paths all departure times on a discrete time dynamic network: the CLASSIC (direct extensions static path algorithms), decreasing order (DOT) algorithm and STEN based an explicit network expansion. All are analyzed under same type compact space-time expansion networks, implemented in flexible extendable framework using object-oriented programming (OOP) concepts. Two randomly generated networks...