Modeling driver behavior in the dilemma zone based on stochastic model predictive control

Area of Science:

• Traffic Safety
• Human-Computer Interaction
• Mathematical Modeling

Background:

• Driver behavior is crucial for traffic safety, particularly in dilemma zones at signalized intersections.
• Accurate driver behavior models can enhance traffic signal control and reduce accidents.
• Understanding driver-vehicle-environment systems is key to mitigating risks in dilemma zones.

Purpose of the Study:

• To develop a mathematical model for driver behavior in dilemma zones.
• To incorporate dynamic human cognition and execution characteristics into driver behavior modeling.
• To provide a feasible solution for more accurate driver behavior modeling and improved system understanding.

Main Methods:

• Utilized stochastic model predictive control (SMPC) for driver behavior modeling.
• Developed a framework including perception, decision-making, and operation modules.
• Simulated driver behavior in dilemma zones using CarSim software for verification.

Main Results:

• The proposed SMPC-based model accurately simulates driver decision-making in dilemma zones.
• The model effectively captures the dynamic characteristics of human cognition and execution.
• CarSim simulations validated the model's ability to reflect vehicle motion and dynamics.

Conclusions:

• The SMPC-based driver behavior model offers a robust approach to understanding and predicting driver actions in dilemma zones.
• This research contributes to improved traffic safety by providing a more accurate driver behavior model.
• The findings enhance the comprehension of complex interactions within driver-vehicle-environment systems.