LSTM-based Prediction of Excavating Resistive Force Using Bucket Trajectory Shape

Unmanned construction machine has become a promising technology in the civil engineering field. The is still less efficient, especially excavating soil, than manned because unmanned works carefully to avoid potential hazards due unpredicted soil deformation as well excavation force. This research aims predict time-series resistive force of bucket during based on Recurrent Neural Network. First, we experimentally obtain for various trajectories. In experiment, follows pre-determined trajectories with four different elliptical arcs, and acting measured. prediction model then elaborated by indexing validation confirms that aspect ratio ellipse an effective index accurately predicting