A 3-D Point-Cloud-Based Automatic Reclaiming Strategy for Portal Scrapers in Complex Stockpile Geometries

Efficient and intelligent operation of raw-material yards is essential for reducing energy consumption in iron and steel production. However, fully automated reclaiming of complex and irregular stockpiles remains challenging due to variable stockpile geometries and uncertain material surfaces. To address this issue, this paper proposes a novel fully automatic reclaiming strategy for portal scraper reclaimers based on a three-dimensional (3-D) point cloud model of the stockpile. The method first establishes a kinematic reclaiming model of the scraper and constructs a TP-PA (travel position-pitch angle) line chart that relates the scraper boom geometry to the stockpile surface. A convex reclaiming-interval identification algorithm is then developed to determine landing points and angles for both flattening and reclaiming operations. Furthermore, a real-time interval fusion mechanism is designed to adapt the reclaiming trajectory as the pile shape evolves. Experimental verification with multiple irregular pile geometries demonstrates that the proposed method enables continuous, geometry-aware reclaiming with minimal idle motion, resulting in improved energy efficiency and operational stability. The proposed strategy is practical for deployment in intelligent, high-efficiency material-handling systems.