Understanding Traffic Dynamics: Microscopic Modeling
Overview
Microscopic traffic flow modeling is essential for linking individual driving behaviors with system-level traffic performance. Traffic stability, safety, congestion, fundamental diagrams, shockwave propagation, and network-level hysteresis all emerge from local interactions involving car-following, lane-changing, perception, reaction, and heterogeneous driving decisions. Microscopic models therefore provide the foundation for explaining how local driving mechanisms shape aggregate traffic dynamics.
Together, my team’s studies establish a multi-scale account of traffic dynamics: from perception and heterogeneous driving behavior to traffic stability and network-level traffic phenomena.
- In our early work, we proposed a visual imaging car-following model, showing that drivers’ visual perception of the preceding vehicle can better explain following behavior and traffic stability, especially under different vehicle-type conditions[3].
- We then examined vehicle-type-dependent car-following heterogeneity from both microscopic and macroscopic viewpoints, demonstrating that larger vehicles lead drivers to maintain lower-risk and larger-spacing behavior[2].
- To connect microscopic behavior with aggregate traffic flow characteristics, we developed a flexible traffic stream model that links reaction time, calmness, and sensitivity with macroscopic fundamental diagrams and shockwave dynamics[1].
- We also derived an anisotropic continuum model directly from car-following behavior and showed its capability to reproduce real congestion waves[4].
- At the network scale, we identified a figure-eight hysteresis pattern in the macroscopic fundamental diagram and traced its microscopic causes to location-specific flow–occupancy states and different lane-changing rates, demonstrating how local maneuvers and spatial traffic heterogeneity shape aggregate network dynamics[5].
- We expanded these foundations to mixed automated–human traffic and priority intersections, modeling how vehicle composition, communication range, human perception and trust, major-road demand, and headway distributions determine car-following stability, fuel use, and minor-road capacity[6,7].
- At freeway merging areas, trajectory data support both behavioral and safety analysis: clustering reveals cautious, moderate, and aggressive driving styles, while joint Tobit and survival models characterize conflict severity and duration[8,9]. Lane-changing simulations further locate effective upstream warning points under different traffic flows[10].
- Two reviews organize the empirical basis of modern traffic-flow modeling: one maps how trajectory data have reshaped microscopic, mesoscopic, and macroscopic studies[11]; the other classifies uses and limitations of the high-fidelity NGSIM datasets and identifies priorities for future data collection[12].
Publications
[1]A Flexible Traffic Stream Model and Its Three Representations of Traffic Flow
[2]Empirical Validation of Vehicle Type-Dependent Car-Following Heterogeneity from Micro- and Macro-Viewpoints
[3]A New Car-Following Model from the Perspective of Visual Imaging
[4]An Anisotropic Continuum Model and Its Calibration with an Improved Monkey Algorithm
[5]A Figure-Eight Hysteresis Pattern in Macroscopic Fundamental Diagrams and Its Microscopic Causes
[6]Modeling the Mixed Traffic Capacity of Minor Roads at a Priority Intersection
[7]A Mixed Traffic Car-Following Behavior Model
[8]Analysis of Vehicle Driving Styles at Freeway Merging Areas Using Trajectory Data
[9]Traffic Conflict Analysis at Freeway Interchange Merging Areas: A Joint Model Based on Tobit and Survival Analysis Model
[10]Optimal Location of Lane-Changing Warning Point in a Two-Lane Road Considering Different Traffic Flows
[11]Trajectory Data-Based Traffic Flow Studies: A Revisit
[12]Research Based on High-Fidelity NGSIM Vehicle Trajectory Datasets: A Review