Adaptive Trajectory Control and Obstacle Avoidance in Autonomous Vehicles Using MPC and RL-Enhanced Decision-Making

The proposed method incorporates the Rapidly Exploring Random Tree (RRT) algorithm, Model Predictive Control (MPC), and Reinforcement Learning (RL) for adaptive trajectory control and obstacle avoidance in autonomous vehicles. Global path planning is attained with the help of the RRT algorithm, while potential field methods address the obstacle avoidance. The precise path following and trajectory control are ensured with the MPC framework, and the adaptive decision-making capabilities of RL enable it to respond to dynamic conditions. The simulation results are validated to demonstrate the effectiveness of the integrated control strategies in dynamic traffic scenarios. The results highlight the improvements in trajectory tracking, decision-making in uncertain conditions, and overall navigation performance.