How do macroscopic traffic flow parameters affect time spent in conflict on freeways? A comprehensive analysis using...

Area of Science:

• Traffic Engineering
• Transportation Safety
• Statistical Modeling

Background:

• Time Spent in Conflict (TSC) is a key metric for freeway safety.
• Understanding the relationship between traffic flow and TSC is vital for effective safety analysis.
• Existing models may not fully account for data censoring and heterogeneity in TSC.

Purpose of the Study:

• To develop a robust model for Time Spent in Conflict (TSC) that addresses data censoring and heterogeneity.
• To investigate the influence of macroscopic traffic flow parameters on TSC.
• To provide a theoretical basis for enhanced freeway safety assessment.

Main Methods:

• Utilized hazard-based duration models, specifically the Kaplan-Meier (K-M) model and the random effect Accelerated Failure Time (REAFT) model.
• The K-M model was used to analyze temporal and spatial heterogeneity of TSC.
• The REAFT model (specifically, the random effect Weibull AFT model) was employed to fit TSC distribution and quantify parameter impacts.

Main Results:

• TSC demonstrates temporal heterogeneity across months.
• The random effect Weibull AFT model provided the best fit for TSC data.
• Traffic density was found to have the most significant positive impact on TSC (24.73%), while flow had a minimal negative impact (-0.6%).

Conclusions:

• The study successfully modeled TSC, accounting for censoring and heterogeneity.
• Macroscopic traffic flow parameters, particularly density, significantly influence TSC.
• Findings support the development of advanced freeway safety systems for accurate crash risk prediction.