Adaptive Control Approach for Optimized Lane Keeping in Autonomous Vehicles

Abstract

The section describes a study on the application of Adaptive Model Predictive Control (AMPC) for lane tracking in autonomous vehicles. The study utilizes a 3 Degree-of-Freedom bicycle model and incorporates dynamic adjustments through sensor fusion to enhance the steering control system, which is crucial for autonomous vehicles. The research focuses on the systematic tuning of control weights and real-time adjustments to demonstrate AMPC's effectiveness in various driving scenarios. Simulation results validate the accuracy, safety, and reliability of AMPC, with ongoing objectives including further development and evaluation of sensor fusion, particularly using depth cameras. The use of Unreal Engine 3D simulation highlights the potential for real-time implementation of AMPC in autonomous driving and lane following. Continuous advancements in control algorithms and sensor technologies are expected to improve AMPC performance further.