Design of multiple helix attractor including operation progress time for semi-autonomous area excavation

Abstract

Unmanned excavation or remote excavation require an autonomous control design. In addition, a semi-autonomous system that can be operated by human intervention in emergency situations is necessary. We have designed an autonomous system using an nonlinear dynamics with an orbit attractor and embedded multiple orbits by extending the dimension of the dynamics. In this paper, we extend the conventional method to design a semi-autonomous system for area excavation. We design an autonomous system whose attractor is a cylindrical surface by helixing trajectories in the extended dimension using index value that represents the operation progress and further multiple helixing. A semi-autonomous system that changes its trajectory by human operation is designed by estimating state variables including index values using an Extended Kalman Filter based on human operation input. The effectiveness of the proposed method is evaluated through experiments using a leader-follower system.