Effect of Sampling Time on MPC for Emergency Single-Lane Change Collision Avoidance: a Case Study of 1:10 Model RC Car

This study investigates the effect of sampling time on Model Predictive Control (MPC) for emergency single-lane-change maneuvers using a 1:10 scale RC car model. Previous implementations showed collisions at a critical cornering speed of 0.5 m/s but could avoid high-speed lane changes up to 2.0 m/s with conventional sampling times (T<inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">s</inf> = 0.1s). Therefore, this study systematically evaluated the effect of sampling times ranging from 0.2s to 0.01s on steering angles, yaw angles, lateral displacements, and computational costs. Results indicated that optimizing the sampling time to T<inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">s</inf> = 0.02s significantly improved control performance, achieving a 60.3% reduction in maximum lateral error and successful collision avoidance at low speeds while maintaining effectiveness at higher velocities (1.0-2.0 m/s). Post-horizon analysis confirmed sustained stability beyond the reference trajectory period. Although frequent sampling times required significant computational resources, the improved safety and precision of control made this a worthwhile trade-off. These findings provide valuable insights for refining the emergency maneuvering of autonomous vehicles across various speeds.