Integrating LiDAR Sensor Data into Microsimulation Model Calibration for Proactive Safety Analysis.

Area of Science:

• Transportation Engineering
• Traffic Safety Analysis
• Data Science

Background:

• Vehicle trajectory data are crucial for calibrating traffic microsimulation models.
• Light Detection and Ranging (LiDAR) provides high-resolution 3D environmental data suitable for real-time applications like autonomous vehicles.
• LiDAR offers advantages over traditional methods due to its speed, accuracy, and performance in various conditions.

Purpose of the Study:

• To develop a framework for integrating LiDAR sensor data into simulation models for proactive safety analysis.
• To calibrate microsimulation models using LiDAR-extracted vehicle trajectory data.
• To assess the effectiveness of calibrated models in predicting traffic safety events.

Main Methods:

• Vehicle trajectory data were extracted from LiDAR point clouds at six urban intersections.
• PTV VISSIM software and the Wiedemann 1999 car-following and lane-changing model were used.
• The Directed Brute Force method calibrated model parameters, achieving 92.7% average accuracy.
• Rear-end conflicts from calibrated models were combined with historical crash data and analyzed using a Negative Binomial model.

Main Results:

• The calibrated microsimulation models accurately replicated observed traffic scenarios.
• Rear-end conflict counts were statistically significant predictors of observed rear-end crash frequencies across all intersections.
• The study demonstrated the feasibility of using LiDAR data for detailed traffic analysis.

Conclusions:

• The integrated framework provides a robust method for proactive safety evaluation using LiDAR data and microsimulation.
• This approach enhances the design of safer transportation systems and traffic control strategies.
• The findings are valuable for transportation professionals seeking to predict and mitigate traffic congestion and safety issues.