Spatio-Temporal Synchronization of Cross Section Based Sensors for High Precision Microscopic Traffic Data...

Area of Science:

• Transportation Engineering
• Traffic Flow Theory
• Sensor Fusion

Background:

• Intelligent Transportation Systems (ITS) demand high-resolution traffic data beyond aggregated parameters.
• Current infrastructure sensors capture individual vehicle data at cross-sections but struggle with longer-range tracking.
• Accurate, continuous vehicle location estimation is crucial for advanced traffic analysis.

Purpose of the Study:

• To develop and validate methods for fusing data from multiple cross-section sensors to reconstruct individual vehicle trajectories.
• To investigate the impact of sensor accuracy on the derived traffic data.
• To enable the derivation of non-linear vehicle trajectories without requiring precise sensor synchronization.

Main Methods:

• Utilizing datasets of vehicle timestamps and speeds as input.
• Implementing a sensor offset estimation algorithm with simultaneous vehicle registration.
• Employing quintic Bézier curves for data fusion and trajectory reconstruction.
• Analyzing the sensitivity of results to individual sensor accuracy.

Main Results:

• A closed-form algorithm for sensor offset estimation and vehicle registration was developed.
• Quintic Bézier curves effectively fused datasets to reconstruct microscopic traffic data.
• The method demonstrated the derivation of non-linear vehicle trajectories.
• The dependency of trajectory reconstruction on sensor accuracy was thoroughly investigated.

Conclusions:

• The proposed data fusion method enhances the utility of existing cross-section sensors for ITS.
• Accurate reconstruction of individual vehicle trajectories is achievable even with imperfect sensor synchronization.
• This approach facilitates more detailed microscopic traffic analysis for future ITS applications.