Digital twin-based excavation trajectory generation of Uncrewed excavators for autonomous mining

Abstract

As modern large-capacity excavators, cable shovels provide payload capacities of more than 100-tons per scoop. The presence of a high dynamic payload combined with the self-weight may result in severe stress loading of the whole shovel front-end components during excavation. Structural performance analysis based on numerical simulations typically demands a prohibitively heavy computation, so conventional optimization-based excavation trajectory generation methods generally do not consider structural performance constraints, which may cause sudden component failure or unanticipated costly downtime. Herein, a shape–performance integrated digital twin was constructed to analyze the shape and structural performance of an uncrewed cable shovel (UCS) in real time, which was further used to support optimal excavation trajectory generation. The proposed procedure presents salient features that the time-consuming structural performance response was taken as an additional constraint, which could effectively prevent the sudden structural failures. The effectiveness and applicability were demonstrated via field tests with a UCS prototype (1:7 scale model of the WK-55 type).