On-line trajectory planning for autonomous robotic excavation based on force/torque sensor measurements

This paper describes the authors' initial investigation into on-line trajectory planning for autonomous robotic excavation. A dynamic model for arm/environment interaction is proposed. Excavation trajectories are generated using a trapezoidal velocity profile under the maximum force/torque constraints. A scheme for real-time identification of dynamic arm/environment interaction parameters is implemented with a first-order algorithm of Saridis-Stein stochastic approximation. The result of a simulation study is presented to illustrate the proposed trajectory planning approach.<>