Under the ongoing “Smart Mine” strategy, intelligent and unmanned mining operations have become a key driver of industry transformation. As critical equipment in open-pit mining, enabling intelligent collaboration between electric shovels and autonomous trucks is essential for advancing unmanned operations. In complex operating environments, however, modeling uncertainties and dynamic excavation resistance can degrade the trajectory-tracking accuracy and robustness of electric shovels, posing significant challenges to precise and stable excavation. To address these issues, three trajectory-tracking strategies—sliding mode control (SMC), fuzzy adaptive sliding mode control (FA-SMC), and RBF adaptive neural network sliding mode control (RBF-ANN-SMC)—are developed. A digital twin platform integrating virtual–physical mapping with closed-loop control is established to support adaptive trajectory tracking under irregular and unstructured pile conditions. Using this platform, simulation analyses are conducted by coupling a multibody dynamics model with a discrete element model of the excavation material, and a scaled prototype is built to perform physical excavation experiments under different material pile-surface conditions. The results demonstrate that the RBF-ANN-SMC-based platform achieves improved tracking accuracy and robustness under complex conditions, providing a practical reference for enhancing the operational efficiency and automation level of electric shovels.
Zhao et al. (2026) studied this question.