Traffic conflict identification method on curved road based on Frenet coordinate system.

Area of Science:

• Traffic Engineering
• Road Safety Analysis
• Vehicle Dynamics

Background:

• Traditional Cartesian coordinate systems struggle with defining and identifying traffic conflicts on curved roads.
• Existing methods often lead to missed or incorrect judgments of rear-end and lane-change conflicts in non-straight road sections.

Purpose of the Study:

• To propose a novel method for accurately identifying traffic conflicts on curved road segments.
• To improve the reliability of Time to Collision (TTC) and derivative indicators for conflict analysis in complex road geometries.

Main Methods:

• Establishment of a Frenet coordinate system based on the road centerline to simplify trajectory and conflict calculations.
• Determination of vehicle states within the Frenet system for calculating rear-end and lane-change conflicts using TTC.
• Validation through video data analysis of a lane-switch work zone on a curved highway section.

Main Results:

• The new method significantly reduces missed judgments of serious rear-end conflicts, particularly at curve-straight road junctions.
• It also decreases erroneous identifications of serious lane-change conflicts on curved roads.
• Identified conflicts showed high deceleration values, indicating the method's effectiveness in detecting hazardous situations.

Conclusions:

• The proposed method, integrating Frenet coordinates, vehicle state determination, and TTC, enhances traffic conflict identification accuracy on curved roads.
• This approach expands traffic conflict analysis from straight to full-line road alignments, improving overall road safety assessment.
• The findings highlight the method's potential to better identify and mitigate risks associated with curved road driving.