Self-Aware Risk Perception and Adaptive Safety Control for Autonomous Mining Trucks

Ensuring the driving safety of autonomous mining trucks (AMT) in unstructured mining environment remains challenging due to irregular road boundaries, unmarked lanes, and irregularly shaped obstacles. This study proposes an autonomous risk perception and safety control method designed for such environment. A spatial risk field and a collision risk field are constructed to autonomously quantify multi-source risk, and an adaptive control strategy is designed based on risk levels, forming an integrated framework from self-aware risk assessment to adaptive safety control. The proposed method is validated using a mining-specific simulator. In normal obstacle encounter scenarios, the AMT maintains an effective safe distance of over 15 m from obstacles. In abnormal deviation scenarios, it can promptly identify risk and stop effectively. Experimental results demonstrate that the proposed method can accurately assess risk and adaptively control the AMT speed, effectively ensuring the driving safety of AMT in unstructured mining environment.