The objective in this paper is to control a robot as it transitions from a non-contact to a contact state with an unactuated viscoelastic mass-spring system such that the mass-spring is regulated to a desired final position. A nonlinear Hunt-Crossley model, which is physically consistent with the real behavior of the system at contact, is used to represent the viscoelastic contact dynamics. A Neural Network feedforward term is used in the controller to estimate the environment uncertainties, which are not linear-in-parameters. The NN Lyapunov based controller is shown to guarantee uniformly ultimately bounded regulation of the system despite parametric and nonparametric uncertainties in the robot and the viscoelastic environment respectively. The proposed controller only depends on the position and velocity terms, and hence, obviates the need for measuring the impact force and acceleration. Further, the controller is continuous, and can be used for both non-contact and contact conditions.
- Dynamic Systems and Control Division
Neural Network Control of a Robot Interacting With an Uncertain Hunt-Crossley Viscoelastic Environment
Bhasin, S, Dupree, K, Patre, PM, & Dixon, WE. "Neural Network Control of a Robot Interacting With an Uncertain Hunt-Crossley Viscoelastic Environment." Proceedings of the ASME 2008 Dynamic Systems and Control Conference. ASME 2008 Dynamic Systems and Control Conference, Parts A and B. Ann Arbor, Michigan, USA. October 20–22, 2008. pp. 875-882. ASME. https://doi.org/10.1115/DSCC2008-2222
Download citation file: