Cooperative Intersection with Misperception in Partially Connected and Automated Traffic.

Area of Science:

• Traffic Engineering
• Intelligent Transportation Systems
• Vehicle Dynamics

Background:

• Connected and automated vehicles (CAVs) promise enhanced traffic efficiency and safety through intersection cooperation.
• Sensor perceptual errors and the coexistence of CAVs with conventional vehicles present challenges to traffic flow.
• Dynamic perceptual errors and time headway significantly impact traffic performance in cooperative intersection scenarios.

Purpose of the Study:

• To develop a simulation model for analyzing the impact of perceptual errors and time headway on traffic performance at cooperative intersections.
• To propose a data fusion scheme for improving perception accuracy and safety in mixed CAV and conventional vehicle traffic.
• To evaluate the trade-offs between traffic efficiency, safety, energy consumption, and congestion under varying CAV penetration rates.

Main Methods:

• Extended the Intelligent Driver Model (IDM) with the Ornstein-Uhlenbeck process to dynamically model perceptual errors.
• Implemented a longitudinal control model for vehicle dynamics, platooning, and reduced deceleration.
• Proposed a data fusion scheme using Differential Global Positioning System (DGPS) data and Kalman filtering to interpolate sensor data for accurate perception.

Main Results:

• Perceptual error and time headway significantly affect crash rates, energy consumption, and congestion at cooperative intersections.
• Larger vehicle intervals reduce accidents but excessive time headway can decrease traffic efficiency and energy conversion.
• The proposed data fusion scheme outperforms independent on-board perception, improving traffic flow, reducing congestion, and enhancing energy efficiency across different CAV penetration rates.

Conclusions:

• A dynamic simulation model effectively captures the influence of perceptual errors and time headway on cooperative intersection traffic.
• Data fusion is crucial for mitigating perceptual errors, enhancing safety, and optimizing traffic flow in mixed CAV environments.
• The study highlights a critical trade-off between safety and efficiency, with the proposed data fusion offering a balanced solution for future transportation systems.