Field measurement analysis to validate lane-changing behavior in a cellular automaton model

Area of Science:

• Traffic Engineering
• Transportation Science
• Computational Transportation Science

Background:

• Microscopic traffic simulation models are crucial for understanding and predicting traffic flow.
• Validation of these models with real-world data is essential for their accuracy and reliability.
• Understanding lane-changing behavior is key to improving traffic efficiency.

Purpose of the Study:

• To validate microscopic traffic simulation models using field measurement data from a Japanese expressway.
• To analyze and confirm common features of traffic flow, such as the fundamental diagram and lane-usage ratio.
• To investigate the specifics of lane-changing behavior, including its asymmetry and the criteria influencing it.

Main Methods:

• Analysis of field measurement data collected from a Japanese expressway.
• Validation of microscopic simulation models against empirical traffic data.
• Examination of the fundamental diagram (flux vs. density) and lane-usage ratio vs. density diagrams.
• Investigation into the factors influencing lane-changing decisions.

Main Results:

• Confirmed common features of the fundamental diagram and lane-usage ratio, consistent with previous research.
• Identified asymmetric lane-changing behavior: changing from a slow to a fast lane differs from changing from a fast to a slow lane.
• Determined that lane-changing incentives depend on gap size: small gaps prioritize velocities/relative velocities, while large gaps also consider distances to preceding vehicles.

Conclusions:

• The study successfully validated microscopic simulation models using real-world expressway data.
• Lane-changing behavior is more complex than often modeled, exhibiting asymmetry and varying incentive criteria based on traffic conditions.
• Findings provide a more refined understanding of lane-changing dynamics, crucial for enhancing traffic simulation accuracy and traffic management strategies.