Navigating Narrow Margins: A Behavior-Based Control Approach for Autonomous Mining Vehicles in Confined Underground Environments

Abstract

Autonomous navigation in underground mining poses a unique set of challenges, from GPS unavailability and high dust levels that compromise visual sensors to feature-poor environments that complicate localization and narrow tunnels that restrict vehicle movement. To address these issues, this paper presents a novel behavior-based control approach, integrating wall-following for lateral stability as the vehicle progresses toward designated positions. The path to these targets is generated by an A* algorithm, ensuring efficient route planning within confined spaces. For localization, an Extended Kalman Filter (EKF) fuses data from wheel odometry and an Inertial Measurement Unit (IMU), providing robust state estimation in the absence of GPS. The proposed system leverages a four-wheel steering mechanism with negative-phase control and is equipped with 3D LiDAR, ultrasonic sensors, wheel encoders, and an IMU for enhanced situational awareness and control. Simulation results validate the system’s ability to achieve precise navigation in challenging underground environments, even within tunnels that allow minimal clearance.