Unmanned excavation using open-pit excavators: algorithmic control and bucket path modelling

The paper examines contemporary approaches to unmanned excavation using open-pit excavators, with a focus on reducing the human factor through data-driven and digital-model-based control. The objective is to develop and justify a methodological approach to algorithmic control of the digging and loading cycle. The method involves formalizing bucket path control as a multi-criteria optimization problem subject to constraints on the drive loads, face geometry, and safety requirements, combined with the system simulation based on the sensor data and a digital face model. It is shown that a closed-loop control system with feedback on loads and position of the working equipment stabilizes the cycle parameters, reduces the operating variability by 15–25%, and limits the peak loads by 10–18%. Methods are additionally considered for adapting the algorithms to changing face conditions, including the rock fragmentation analysis and telemetry data. It is established that introduction of a virtual loading zone and integration with the in-pit fleet management system improve safety and quality of loading. The results demonstrate that transition to the algorithmic control based on digital models creates prerequisites for conditionmonitored maintenance and further digital transformation of mining operations.