The simulation dataset of each model for the circle antipode experiment

Area of Science:

• Pedestrian dynamics and crowd simulation.
• Agent-based modeling.
• Behavioral science.

Background:

• Existing pedestrian simulation models struggle to accurately capture detour behavior.
• The circle antipode experiment is a standard test for evaluating pedestrian movement models.
• Understanding pedestrian avoidance is crucial for safe crowd management.

Purpose of the Study:

• To provide a dataset of trajectory and video data from simulations of the circle antipode experiment.
• To highlight discrepancies in pedestrian detour behavior simulation among different models.
• To offer comparative data for enhancing pedestrian simulation models.

Main Methods:

• Simulating the circle antipode experiment using the traditional social force model, Voronoi-based detour social force model, and double-layer detour decision model.
• Recording pedestrian coordinates at each simulation step to generate trajectory data.
• Capturing each frame during simulation to create video data.

Main Results:

• The dataset reveals distinct differences in simulated pedestrian detour paths across the evaluated models.
• Trajectory and video data offer a clear visualization of model limitations in replicating avoidance.
• The generated data serves as a benchmark for model validation and improvement.

Conclusions:

• The provided dataset is valuable for identifying and addressing gaps in pedestrian simulation models.
• This data can guide the modification of existing models to better represent detour behavior.
• The findings support the development of more realistic pedestrian flow simulations.