Vector Field-based Autonomous Navigation in a Tunnel-like Environment

Licheng Feng, Jianjun Bao, Binghao Li, Liao Wu

Abstract: Tunnel-like environments are often extremely long and large-scale, particularly narrow and confined. This paper investigates the problem of autonomously guiding a robot through a tunnel-like environment. To address this challenge, we initially create a skeleton-based reference curve to compute the normal and tangential direction of the tunnel-like environment. Then we create a barrier function to delineate the impenetrable nature of the environment’s boundary. Subsequently, we employ the principle of vector field to design a novel navigation law that guarantees the continuous advancement of the robot within the tunnel meanwhile simultaneously maintaining a predefined safety margin from the bounds of the environment. Conducted simulation experiments validate the effectiveness of the proposed vector field-based strategy for safely navigating a robot through tunnel-like environments.