Unveiling Multistability in Urban Traffic Through Percolation Theory and Network Analysis

Area of Science:

• Complex systems science
• Urban planning
• Network science

Background:

• Traffic congestion is a persistent urban challenge.
• The multistability of urban traffic flow is poorly understood.
• Existing models do not fully capture complex traffic dynamics.

Purpose of the Study:

• To empirically investigate multistability in urban traffic flow.
• To identify key network features influencing traffic states.
• To provide a quantitative basis for traffic management.

Main Methods:

• Analysis of high-resolution traffic data from four Chinese cities.
• Application of percolation theory to characterize traffic congestion.
• Use of data-driven clustering and classification techniques.

Main Results:

• Observed bimodal/multimodal distributions of congestion rate (f) and largest functional cluster size (G), confirming multistability.
• Identified high betweenness centrality road segments as critical congestion points.
• Top 1% of topologically important roads accurately predict traffic states and dynamics.

Conclusions:

• First large-scale empirical evidence of multistability in urban traffic networks.
• Topological importance of road segments is a key predictor of traffic behavior.
• Findings support improved traffic forecasting and management strategies.