Enhancing Motivational Driver Model Using Model Predictive Control for Autonomous Decision-Making

This study investigates the integration of Model Predictive Control (MPC) with a cognitive human driver behavior model to enhance autonomous vehicle decision-making in dynamic traffic environments. Building on a foundational driver model informed by psychological theories, the research refines decisionmaking processes by leveraging MPC to anticipate and account for potential traffic events. A three-degree-of-freedom vehicle dynamics model, combined with sliding-mode and proportional controllers for lateral and longitudinal motions, respectively, is developed to facilitate simulations. The simulation results reveal that the proposed MPC-enhanced driver model significantly improves the ego-vehicle's ability to respond to dynamic traffic scenarios, such as a sudden lane change by a moving obstacle in front of the vehicle. Our method avoids collisions during critical maneuvers while achieving smoother and more stable trajectories compared to the conventional driver model.