Neural network-based teleoperation using Smith predictors

The introduction of communication channel tune delay and environment dynamic uncertainties create a significant challenge in the design of stable transparent bilateral teleoperation controllers. An early control methodology for time delayed systems, which is applicable to teleoperation systems is the use of Smith predictors. Recently a few Smith predictor based teleoperation control architectures have been proposed for 2-channel teleoperation systems in which the linear dynamics of the slave or environment are mapped at the master. This paper discusses the effectiveness of this control structure for 2-channel force-position teleoperation when applied to the nonlinear time varying dynamics of slave and environment. The proposed nonlinear predictive controller and its variations use neural networks to online estimate the dynamics of the slave and environment allowing replication of the environment contact force at the master using a similar network. The performance of the proposed architectures are evaluated on a teleoperation test-bed consisting of two planar twin-pantograph haptic devices.