Multilevel Health Monitoring of Autonomous Mining Trucks Under Complex Time-Varying Conditions

The deployment of autonomous mining trucks (AMTs) has become a major trend in mining operations, driven by the need for enhanced safety and operational efficiency. This study addresses the challenge of monitoring AMT motion status under time-varying and complex working conditions to reduce accident risks. A multilevel health monitoring approach is proposed, incorporating a weighted clustering method to classify working conditions and improve the assessment of key health indicators. The proposed method refines the evaluation of working conditions by applying a targeted clustering approach, which enables more precise health monitoring tailored to each condition category. Experimental validation using data from real-world open-pit mines demonstrates that this approach significantly improves the accuracy and timeliness of fault detection, thereby enhancing overall operational safety and reliability.