Adaptive Vehicle Path Planning in Alleyway Using Looming ODG-MPC

초록

Last-mile driving environments, such as alleyways, present significant challenges for autonomous vehicles due to their narrow roads and the presence of various obstacles. Achieving fully autonomous driving in these settings requires both precise maneuvering and effective collision avoidance. In this paper, we propose a novel looming obstacle dependent Gaussian-based model predictive control (Looming ODG-MPC) algorithm for adaptive path planning in alleyways. Our approach introduces a two-dimensional obstacle dependent Gaussian potential field (2D ODG-PF) to generate obstacle avoidance paths that consider both the shape and heading angle of obstacles. By incorporating collision probability calculations and adaptive safe driving speeds, our method enhances tracking performance in confined spaces. Additionally, the T-S fuzzy approach is employed to model time-varying changes in vehicle velocity, addressing the linear parameter-varying nature of the system. To validate the effectiveness of the proposed algorithm, we conducted comparative experiments against existing state-of-the-art methods. The results demonstrate that our algorithm maintains a safer distance from obstacles, reduces unnecessary steering, and improves ride comfort, outperforming the existing algorithms in complex and narrow environments.

키워드